Liquid immersion cooling platform

Abstract

A two-phase liquid immersion cooling system is described in which a heat-generating computer component evaporates a dielectric fluid in its liquid phase. The dielectric vapor is then condensed back to the liquid phase for cooling computer components. Using a pressure controlled vessel and a pressure controller, the disclosed system can operate at sub-ambient pressure. By controlling the pressure at which the system is operated, a user may affect the temperature at which the dielectric fluid evaporates and thereby achieve improved performance of a given computer component. A self-healing computing system may be created using a robotic arm and a slot-based computing component. A two-phase liquid immersion cooling system is described in which a heat-generating computer component evaporates a dielectric fluid in its liquid phase. The dielectric vapor is then condensed back to the liquid phase for cooling computer components. Using a pressure controlled vessel and a pressure controller, the disclosed system can operate at sub-ambient pressure.

Description

[0001] This application is a divisional application of an application with an application date of November 11, 2019, application number 201980089348.3 (international application number PCT / US2019 / 060759), and invention name “Liquid Immersion Cooling Platform”. Technical Field

[0002] The present invention relates to liquid immersion cooled computing systems, ie, computing systems utilizing liquid immersion cooling with pressure and / or vapor management. Background Art

[0003] Conventional computing and / or server systems utilize air to cool various components. Conventional liquid or water-cooled computers utilize flowing liquid to draw heat from computer components, but avoid direct contact between the computer components and the liquid itself. The development of non-conductive and / or dielectric fluids has enabled the use of immersion cooling, in which computer components and other electronic devices can be immersed in a dielectric or non-conductive liquid in order to draw heat from the components directly into the liquid. Immersion cooling can be used to reduce the total energy required to cool computer components, and can also reduce the amount of equipment and space required for adequate cooling. Summary of the invention

[0004] In the disclosed embodiments of the invention described below, the use of a steam and pressure management system and a power management system may be utilized alone or in combination to create a significantly improved computer system utilizing liquid immersion cooling.

[0005] Embodiments of the disclosed invention relate to a pressure controlled container that can be used to accommodate a computing system for liquid immersion cooling. In some embodiments, the pressure controlled container contains a sufficient amount of liquid dielectric fluid to fully immerse the heat-generating computer components, and also includes an atmosphere containing a gaseous dielectric fluid. Embodiments also include a condensation system to cool and convert the gaseous dielectric fluid into a liquid dielectric fluid. The disclosed pressure management system allows the disclosed embodiments to operate under vacuum, thereby reducing the temperature at which the dielectric fluid evaporates and the computing system operates. Due to the described improved temperature management system, the disclosed embodiments allow the density and / or computing power of computer components to be increased. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1-2 A schematic diagram of a controlled pressure vessel is shown according to an example embodiment.

[0007] Figure 3 The exterior of an example embodiment of a controlled pressure vessel 110 is shown.

[0008] Figure 4 An example embodiment of a superstructure comprising a plurality of pressure controlled vessels is depicted.

[0009] Figure 5 An exemplary data center embodiment is depicted showing multiple pressure controlled vessels connected to a central power source.

[0010] Figure 6 An exemplary data center embodiment is depicted showing a plurality of controlled pressure vessels connected in series with each other.

[0011] Fig. 7A -D depicts an example embodiment of a cooled computing system with an internal robotic arm, an airlock, and an external robotic arm.

[0012] Fig. 8A -C shows an example embodiment of a rack system.

[0013] Fig. 9A -G shows an example embodiment of a chassis for mounting various components.

[0014] Fig. 10A -F shows an example embodiment of a pressure controlled container.

[0015] Fig.11 An example cooling and steam management system for a controlled pressure vessel is shown.

[0016] Fig. 12A -E shows another embodiment of a container.

[0017] Fig.13 An example of a self-contained container is shown.

[0018] Fig.14 An example of an outer shell of a self-contained container is shown.

[0019] Fig.15A -D shows an example cassette positioned on a platform that can extend out of the container.

[0020] Fig.16 A vapor recovery system according to an example embodiment is shown.

[0021] Fig.17 An example embodiment of a rack power distribution system is shown.

[0022] Fig.18 An example of a heating element for an immersion cooling system is shown according to an exemplary embodiment.

[0023] Fig.19A -B shows a filter including three cores according to an example embodiment.

[0024] Fig. 20A -B shows an example robotic system.

[0025] Fig.21A-B shows an exemplary guide pin mechanism between the chassis and the frame.

[0026] Fig. 22 An example connector with self-alignment features is shown. DETAILED DESCRIPTION

[0027] In the following description, certain details are set forth, such as specific numbers, sizes, arrangements, configurations, components, etc., in order to provide a thorough understanding of the current embodiments disclosed herein. However, it will be apparent to one of ordinary skill in the art that the present disclosure may be practiced without these specific details. In many cases, details regarding such considerations, etc. have been omitted because such details are not necessary to obtain a complete understanding of the present disclosure and are within the skill of one of ordinary skill in the relevant art.

[0028] The devices, components, systems and subsystems of some embodiments disclosed below are described according to trade names. It will be apparent to one of ordinary skill in the art that the present disclosure can be practiced with many similar components, whether or not these components are developed and / or sold under a specific trade name, and that the features and / or limitations associated with a specific trade name component are not necessary for practicing the disclosed invention.

[0029] Dielectric Fluid

[0030] One aspect of immersion cooling is the use of thermally conductive but substantially non-conductive or substantially dielectric fluids. Examples of such fluids include 3M TM Novec TM Some of the series of engineering fluids (including Novec 7100), although the invention is not limited to any specific dielectric fluid. Some immersion fluids generally have a certain boiling point, and it is desirable to operate the computer components of the cooling at this boiling point. All computer components and other aspects of the disclosed system are preferably made of materials that are insoluble when in contact with the dielectric fluid and will not rupture in a pressure-controlled container. In some embodiments, the boiling point of the dielectric fluid at standard atmospheric pressure can be less than about 100°C, or less than about 80°C, or less than about 60°C, or less than about 50°C, or even lower. In some embodiments, the boiling point of the dielectric fluid at standard atmospheric pressure is greater than about 60°C, or greater than about 40°C, or greater than about 30°C, or greater than about 20°C. Some embodiments of the immersion cooling fluid generally have a low vapor pressure. Some embodiments of the immersion cooling fluid are fluorocarbons and / or fluorinated ketones. Some embodiments of the dielectric fluid can have (CF3)2CFCF2OCH3, C4F9OCH3 or CF3CF2CF2CF2OCH3 or a chemical formula similar thereto. Some dielectric fluids include hydrofluoroethers, methoxy-nonafluorobutane.

[0031] Other desired characteristics of immersion cooling fluid include low toxicity, nonflammable and / or low surface tension. In certain embodiments, under the pressure and temperature for liquid immersion cooling, immersion cooling fluid will not damage computer components and / or connectors, wires, cables, seals and / or adhesives associated with computer components substantially. Some dielectric fluids have a dielectric constant ranging from about 1.8 to about 8, and a dielectric strength of about 15 megavolts / meter (MV / m). In certain embodiments, dielectric fluid has a dielectric strength of at least about 5MV / m, or at least about 8MV / m, or at least about 10MV / m, or at least about 12MV / m. In certain embodiments, dielectric fluid has a dielectric strength of at most about 3MV / m, or at most about 5MV / m, or at most about 8MV / m. In disclosed embodiments, any liquid in contact with computer components 170 has a sufficiently high dielectric strength to avoid damaging computer components at intervals and conditions of specific applications.

[0032] Some dielectric fluids have a critical heat flux of at least about 10 W / cm2, or at least about 15 W / cm2, or at least about 18 W / cm2, or at least about 20 W / cm2. Some dielectric fluids have a critical heat flux of at most about 15 W / cm2, or at most about 10 W / cm2, or at most about 8 W / cm2, or at most about 5 W / cm2.

[0033] Figure 1 A schematic diagram of a cooling computing system 110 according to an example embodiment is shown. An embodiment of the disclosed cooling computing system 110 (or computing system, system, container or pressure controlled container, all of which can be used interchangeably) can utilize a liquid dielectric fluid 140 to cool a computer component 170 by immersing the component in a bath of fluid. When electricity passes through the component 170, the component 170 generates heat. As the component 170 heats up, the performance of the component may decrease, or the component may be damaged to the point of failure. It is advantageous to maintain various computing components at a stable and relatively low temperature. In some embodiments, the computer component 170 may be maintained at less than about 80°C, or less than about 70°C, or less than about 65°C, or less than about 60°C, or less than about 55°C. In some embodiments, the computer component 170 may be maintained at greater than about 60°C, or greater than about 50°C, or greater than about 40°C, or greater than about 35°C, or greater than about 30°C. When the computer component 170 heats up, the heat is transferred to the liquid dielectric fluid 140 surrounding the component 170. When the liquid dielectric fluid reaches its boiling point, it will transition from the liquid phase to the vapor phase and rise away from the liquid bath 142. The components 170 in the bath 142 of dielectric fluid can generally be maintained at about the boiling point of the particular dielectric fluid 140 being used.

[0034] When the liquid dielectric fluid is heated to the vaporization point and becomes a gas at a pressure for a given application, bubbles of dielectric vapor will rise out of the liquid bath 142 and rise to the top of the system 110. The vapor is then cooled to the condensation point using the condenser 130. Depending on the configuration of the system 110, the heating and cooling of the dielectric fluid from the liquid phase to the vapor phase and from the vapor phase to the liquid phase can produce Figure 2 Convection shown.

[0035] In some embodiments, when the system is running, the computer component 170 will be completely immersed in the liquid dielectric fluid 140. In other words, the upper portion of the computer component 170 is below the horizontal plane of the dielectric liquid 140. It should be understood that due to the heat from the computer component causing the dielectric fluid to change from the liquid phase to the gas phase, the small bubbles of the dielectric fluid vapor will contact the computer component. Such components will still be considered to be completely immersed in the liquid phase of the dielectric fluid. In some embodiments, the computer component 170 can be immersed in the liquid phase of the dielectric fluid 140. In an exemplary embodiment, if any part of the computer component (including but not limited to any part of the motherboard, chip, server, card, blade, GPU or CPU and / or any peripheral component) is in direct contact with the liquid phase of the dielectric fluid 140, the computer component will be considered to be immersed. In certain embodiments, the computer component 170 can be at least partially immersed in the liquid phase of the dielectric fluid 140. If the computer component 170 is not immersed, but is fully cooled by the dielectric vapor, the computer component will be considered to be at least partially immersed.

[0036] In some existing immersion cooling systems, dielectric fluid must be constantly added to the bath of dielectric fluid as the fluid is continuously vaporized. Failure to add dielectric fluid to bath 142 may cause the level of dielectric fluid in bath 142 to drop until the component is exposed to the gas atmosphere and is not adequately cooled. This may result in performance degradation or damage to component 170.

[0037] In some embodiments, there may be multiple modes of operation that may describe a fluid management system for use with a dielectric fluid in its liquid state. These modes may include: (1) initial filling, which is the process of transferring the dielectric fluid from the storage system into the container; (2) continuous level adjustment, which is the process of adding additional fluid to the container or removing excess fluid from the container; (3) defilling, which is the process of draining the fluid from the container and placing it into the storage system; and (4) operational filtration, which is the process of continuously circulating the fluid through the filtration system to ensure that any particles are removed.

[0038] In some embodiments, the first three liquid management objectives, initial filling, continuous level adjustment, and defilling, can be achieved by the same set of pipes, pumps, and valves. A dedicated tank for storing liquid coolant can be used to store new and excess fluid that is removed and recondensed during the vapor management process. A set of pipes and pumps can be used to bring coolant (or dielectric fluid) from the storage system to the container during filling and level adjustment, and back from the container and into the storage system during defilling operations.

[0039] In some embodiments, the fourth liquid management goal, i.e., operational filtration, can be achieved through a series of skimmers and / or filters. The first stage can be a large particle filter located at the bottom of the container. The purpose of this filter is to prevent particles that are too large to be processed by the next stage from entering the rest of the system. The second stage can be a colinear medium particle filter located in the piping system between the first and third stages. This second stage medium particle filter can use a small cartridge filter to remove particles that are too small to be removed by the first stage filter but still too large to be processed by the third stage filter. The third stage filter can be composed of one or more parallel filters and supports various types of filter configurations. In some embodiments, a specific type of filter will be determined by analyzing the fluid after exposure and operating it using a set of hardware components located in the container environment. Different hardware and / or components may produce different types of particles and chemicals that may need to be filtered to ensure the long-term life and efficiency of the dielectric fluid.

[0040] Stress Management

[0041] In general, the immersion cooling fluid must remain free of dust, water and / or other contaminants. Since the computer components 170 are in direct contact with the immersion cooling fluid 140, tiny contaminants may cause short circuits or damage to the computer components. In addition, the water or water vapor that may contaminate the dielectric fluid may reduce the dielectric properties of the fluid when the fluid is contaminated, including but not limited to the dielectric strength of the fluid. If the dielectric strength of the dielectric fluid is reduced, the computer components may short circuit or otherwise be damaged during operation. A way to reduce contamination is to operate the immersion cooling system in a closure that is maintained slightly above or above atmospheric pressure.

[0042] When the computer components 170 are in operation, the heat generated by the initial use of the computer components causes some of the dielectric liquid 140 to evaporate into a gas. If the immersion cooling system is confined within a substantially closed enclosure, this evaporation will generally increase the pressure of the atmosphere within the enclosure. Pressure relief valves, expansion closures, and / or other techniques may be used to limit the increased pressure and / or maintain the pressure within the enclosure at or just above atmospheric pressure. Maintaining a slight positive pressure within the enclosure may help reduce the penetration of dust, water vapor, or other contaminants into the immersion cooled computing system.

[0043] The current embodiment utilizes a closed controlled pressure vessel 110 (or cooled computing system 110) enclosure to house computing components 170 and immersion cooling equipment within the controlled pressure vessel, as well as associated power supplies, network connections, wiring connections, etc. In contrast to prior models, the controlled pressure vessel 110 can be maintained at least at a slight vacuum, thereby reducing the boiling point of the dielectric fluid 140 to a temperature below its boiling point at standard atmospheric pressure.

[0044] By operating the computing and immersion cooling system under vacuum, the components 170 can be maintained at the reduced low pressure boiling point of the dielectric fluid 140. This has the benefit of increasing cooling, which allows more electricity to pass through various components 170, thereby resulting in the higher performance of the components. By controlling the pressure in the pressure controlled container 110, the boiling point of the dielectric fluid 140 can also be controlled, thereby allowing the same fluid 140 to be used under the conditions of a wider range. Many embodiments benefit from colder temperatures, yet some computer components 170 have an ideal range, and performance is reduced at temperatures below the range. By controlling the pressure in the pressure controlled container 110, the boiling point of the immersion cooling fluid 140 can also be controlled. In certain embodiments, the disclosed pressure management system can be used to dynamically control pressure when the computing system is started, shut down, or responds to other changing conditions, and thereby controls the boiling point of the dielectric fluid 140.

[0045] Except reducing the boiling point of dielectric fluid 140 by operating in pressure controlled container 110 below ambient pressure, computer component 170 itself can be modified so that heat is more effectively transferred from itself to dielectric fluid 140. By increasing the surface area of ​​components 170 (such as chips) exposed to liquid dielectric fluid 140, heat transfer between bath 142 of components 170 and dielectric fluid 140 can be increased. Exemplary devices for increasing surface area can be copper boilers or copper pans, which can adhere to the chips of other computer components 170. In certain embodiments, the adhesive used will be selected based on its heat transfer capability and its solubility in dielectric cooling fluid. Preferred adhesives show high thermal conductivity and low solubility in selected dielectric fluids.

[0046] Figure 1A schematic diagram of an example embodiment of the disclosed computing system is shown. An embodiment of the disclosed system includes a pressure controlled vessel 110 (or a cooled computing system 110), a pressure controller 150, an immersion cooling system including at least a volume of dielectric fluid 140 and a condensing structure 130, and desired computer components 170. The pressure system can be configured to maintain a desired degree of reduced pressure. The pressure controlled vessel 110 can be configured to maintain a negative pressure while still allowing multiple penetrations into the pressure controlled vessel 110 for various connections, including but not limited to power, data, network, cooling water and / or communication systems. Some embodiments utilize sealed and / or marine grade connections. Operating a computer system within a pressure controlled vessel 110 below ambient pressure requires a series of modifications to the system as a whole. These modifications are discussed below, and some of them will be apparent to one of ordinary skill in the art.

[0047] Figure 3 The exterior of an example embodiment of a controlled pressure vessel 110 is shown. In some embodiments, the disclosed controlled pressure vessel 110 is at least about 2 feet tall, or at least about 3 feet tall, or at least about 4 feet tall, or at least about 5 feet tall. In some embodiments, the controlled pressure vessel is at most about 3 feet tall, or at most about 4 feet tall, or at most about 5 feet tall.

[0048] In certain embodiments, the pressure controlled vessel has an internal volume of at least about 100 cubic feet, or at least about 150 cubic feet, or at least about 200 cubic feet, or at least about 250 cubic feet, or at least about 300 cubic feet, or at least about 350 cubic feet, or at least about 400 cubic feet.

[0049] In some embodiments, the pressure controlled container will be configured to accommodate approximately 12 vertical inches of liquid dielectric fluid and approximately 36 vertical inches of dielectric fluid vapor during operation. In certain embodiments, the ratio of liquid volume to gas volume helps to generate convection and guide the gaseous dielectric vapor to a condensation structure that converts the vapor back into a liquid. In some embodiments, the pressure controlled container is configured to accommodate a ratio of a certain volume of liquid dielectric fluid to a certain volume of gaseous dielectric fluid of approximately 1:6 during operation. In other embodiments, the pressure controlled container is configured to accommodate a ratio of a certain volume of liquid dielectric fluid to a certain volume of gaseous dielectric fluid of approximately 1:3, or approximately 1:5, or approximately 1:8, or approximately 1:10, or approximately 1:15 during operation.

[0050] In one example embodiment, the pressure management system may include a pressure controller 150. The pressure controller 150 may be a vacuum source, for example, the pressure controller 150 may be a vacuum pump that may be connected to the pressure controlled vessel 110. In some embodiments, the vacuum pump 150 may be remote, and the vacuum may be transferred to the pressure controlled vessel 110 using tubing and / or piping. In a preferred embodiment, a pressure sensor 180 is housed within the pressure controlled vessel 110 and is used to regulate and / or maintain a desired negative pressure within the pressure controlled vessel 110. In some embodiments, the pressure sensor 180 and / or the pressure regulator 190 may be connected to a processor that monitors the pressure in the pressure controlled vessel 110 using the pressure sensor 180 and regulates the pressure using the pressure regulator 190.

[0051] Some embodiments include an operator protection mechanism. In an example embodiment, the operator protection mechanism can be a locking mechanism that prevents the system from operating if any cover or service panel of the pressure controlled vessel is not in place. In an example embodiment, the operator protection mechanism can include a controller to immediately de-energize the system in the event of an unauthorized tampering with one of the doors or panels of the pressure controlled vessel. In addition to providing life safety functions, the operator protection mechanism can also provide enhanced operational safety functions for deployed items within the container that store sensitive data. By ensuring that the equipment cannot be accessed during normal operation without shutting off power to the system, a high level of assurance can be achieved in the efficiency of the disk protection mechanism. In addition, in some embodiments, the disk protection mechanism can use encryption keys stored at runtime to protect static data on the pressure controlled vessel.

[0052] In some embodiments, in addition to denying unsafe access to the pressure controlled vessel, sensors may also be placed to ensure that the system is operating as designed. The primary sensor package may include a temperature sensor in the vapor space; a temperature sensor in the liquid space; a humidity sensor in the vapor space; and / or a pressure sensor in the vapor space. These sensor readings may be monitored by software and / or a human operator to ensure that the system is operating in a safe and correct manner. In some embodiments, the sensor data will be recorded or analyzed later.

[0053] In some embodiments, additional sensors may be incorporated within the container or superstructure (defined below). Such sensors may include, for example, a FLIR-based thermal imaging camera; a VESDA or other form of aspirating smoke detector; and / or a refrigerant leak detector designed to detect leakage of dielectric fluid into the surrounding environment.

[0054] In some embodiments, the container and / or superstructure may be equipped with indicator lights related to the operating status of the system.

[0055] Although the cooled computing system 110 is sometimes referred to as a pressure-controlled system 110, one of ordinary skill in the art recognizes that many, if not all, of the benefits of the cooled computing system 110 may be achieved without using a "pressure-controlled system."

[0056] Steam Management System

[0057] Liquid immersion cooling systems can operate in different ways. Some can operate by directly continuously cooling the immersion fluid. Others can operate by allowing the liquid to reach its maximum liquid phase temperature and then boil into a vapor phase. Immersion cooling systems that operate by allowing the liquid to evaporate are called two-phase immersion cooling systems. Two-phase immersion cooling systems typically allow the dielectric fluid to boil and / or evaporate, and additional fluid is periodically added to replace fluid lost to the atmosphere.

[0058] The disclosed embodiment utilizes the liquid immersion cooling system contained in the pressure controlled container 110. This has the advantage that the dielectric fluid will not be lost even after the dielectric fluid 140 has been converted into gaseous form. In the closed or substantially closed pressure controlled container 110, the gaseous dielectric fluid can be condensed and added back in the bath 142 of the liquid dielectric fluid 140, and the liquid dielectric fluid is actively used for cooling the computing component 170. The condensation step can be carried out in any convenient way, for example, by making the process water flow through the heat pipe. The condensation structure 130 can include radiator fins and / or similar equipment, which increases the surface area of ​​the condenser, thereby allowing the gaseous dielectric fluid to condense larger and / or faster, and return it to liquid form. In certain embodiments, the process water is at ambient temperature, and is not actively cooled. In other embodiments, the method for freezing process water can be used for evaporative cooling, dry cooling tower and / or other freezing process water known in the art.

[0059] In some embodiments, there may be two interfaces between the pressure controlled container and the external system. The first may be a process water supply interface. This may be a pipe that delivers process water from a facility that provides chilled process water to a distribution manifold on the pressure controlled container. The second may be a process water return interface. This may be a pipe that returns process water to the facility that provides chilled water. After the process water has flowed through the pressure controlled container and associated cooling components, the process water may be returned to the facility. The cooling components may include, for example, a condenser, a condensing coil, and / or a radiator within the container, and a coil that removes heat from the exhaust pipe of any power component (including, for example, a motor, a pump, and / or a utility cabinet). In some embodiments, there may be two interfaces between the super structure and the external system. The interface may be similar to or substantially similar to the two interfaces between the pressure controlled container and the external system.

[0060] In some embodiments, the location of the condensation structure 130 within the controlled pressure vessel 110 can be configured to optimize the flow of the vapor phase dielectric fluid and increase the rate and / or efficiency of condensation. In some embodiments, the geometry of the controlled pressure vessel 110 itself can be controlled to increase the rate and / or efficiency of condensation.

[0061] In one example embodiment, the location of the condensation structure 130 can facilitate and optimize the placement (e.g., by a robot) of the computer component 170 within the container (or the removal of the computer component 170 from the container). For example, the condensation structure 130 can be placed on a side (or sidewall) of the container so that the condensation structure 130 is not located between the lid of the container and the computer component 170. In this way, when the lid is opened, the robot can directly remove the computer 170 without interfering with the condensation structure 130. This arrangement of the condensation structure can simplify the placement and removal of the computer component 170, which can provide significant benefits in the autonomous operation of the container. In an exemplary embodiment, the condensation structure 130 can be located above a shelf within the container.

[0062] like Figure 1-3 As shown, in one exemplary embodiment, the pressure controlled vessel is about 10 feet long, about 4 feet wide, and about 4 feet high. TM The dielectric fluid 140 can create a bath 142 in the pressure controlled vessel 110. This leaves a liquid dielectric fluid layer of about 12 inches deep in the immersion cooling tank at the bottom of the pressure controlled vessel, while most of the volume of the pressure controlled vessel is gaseous. The top plate of the pressure controlled vessel is lower in the middle of the longitudinally extending structure. The top plate and / or lid 120 are angled upward and rise as the sidewalls of the pressure controlled vessel 110 are approached. The condensation structure 130 extends longitudinally on both sides of the pressure controlled vessel 110. The condensation structure 130 in this exemplary embodiment can be about 12 inches wide and about 24 inches high, and basically extends the entire length of the pressure controlled vessel 110. The condensation structure 130 includes a radiator-like material with high surface area fins, which are cooled using flowing process water. Some embodiments may additionally or alternatively include a heat exchanger.

[0063] like Figure 2As shown, the structural arrangement in the pressure controlled vessel 110 guides the convection of the dielectric fluid vapor when it rises from the liquid bath 142 after the dielectric fluid vapor boils. The structural arrangement guides the convection upward to the top plate of the pressure controlled vessel, where the flow is directed to the high surface area of ​​the condensation structure 130 and condenses back into liquid form. The dielectric fluid 140 then flows back into the liquid bath 142. In this way, the total amount of the dielectric fluid 140 can be stored in the closed shell. Using convection to circulate the dielectric fluid vapor allows the disclosed embodiment to operate without a mechanical pump for circulating the dielectric liquid, thereby reducing the total energy use of the disclosed system.

[0064] Certain embodiments may utilize additional tanks and / or storage vessels for dielectric fluid that may be used during startup and / or shutdown of the system in case the pressure controlled vessel must be opened and / or to allow for redundant and robust control of the liquid dielectric fluid level.

[0065] Fig.11 An example cooling and steam management system 600 for a pressure controlled vessel 110 is shown. In this example embodiment, the cooling and steam management system 600 may include a refrigerated process water reservoir 611 that extends through the cooling coil 132 to cause condensation of the dielectric fluid 140. After passing through the cooling coil 132, the process water may advance to the process water return reservoir 612. The cooling and steam management system 600 may also include a tank 614 for steam storage and a tank 615 for dielectric fluid storage. When needed, for example, during system startup and / or shutdown, tanks 614 and 615 may provide dielectric fluid or steam. In an example embodiment, tanks 614 and 615 may be coupled via a condensation structure 616. In the case of an excess steam supply in tank 614, the condensation structure 616 may remove steam and add it as a dielectric fluid to the fluid storage tank 615.

[0066] In some embodiments, during operation, the pressure controlled vessel is maintained at about 3 psi below the ambient atmospheric pressure, which helps to reduce the boiling point of the dielectric fluid, thereby reducing the operating temperature of computer chips and other components. In some embodiments, the pressure controlled vessel 110 is maintained at least about 2 psi below ambient pressure, or at least about 4 psi below ambient pressure, or at least about 6 psi below ambient pressure, or at least about 8 psi below ambient pressure, or at least about 10 psi below ambient pressure.

[0067] In some embodiments, it will be necessary to select components that have a certain degree of tolerance for pressure fluctuations. It is preferred to use components that can withstand a wide range of pressures to allow for manipulation of the coolant boiling point, and also the overall operating temperature of the overall system, by adjusting the operating pressure of the system. Given the operating nature of a two-phase system, standard operating conditions for some embodiments will see a difference between ±4 PSIg. In some cases, such as during rapid startup or shutdown of the system, a difference of three additional PSIg may be experienced. In some embodiments, system-level adjustments can be made to better control these variables and keep them within a more controlled and defined range.

[0068] In some embodiments, computer component 170 operates at least about 3% below ambient pressure, or at least about 5% below ambient pressure, or at least about 10% below ambient pressure, or at least about 15% below ambient pressure, or at least about 20% below ambient pressure, or at least about 25% below ambient pressure, or at least about 30% below ambient pressure.

[0069] In some embodiments, during operation, the pressure controlled vessel is maintained at less than about 750 Torr, or less than about 710 Torr, or less than about 650 Torr, or less than about 550 Torr, or less than about 500 Torr, or less than about 450 Torr, or less than about 400 Torr, or lower. In some embodiments, during operation, the pressure controlled vessel is maintained at greater than about 650 Torr, or greater than about 600 Torr, or greater than about 550 Torr, or greater than about 500 Torr, or greater than about 450 Torr, or greater than about 400 Torr, or greater than about 300 Torr.

[0070] Some embodiments utilize steam washing process and / or initial purification process in order to control the gaseous atmosphere in the pressure controlled container.This process removes part of the gaseous atmosphere from the pressure controlled container, and removes a part of unwanted atmosphere, such as air and water vapor.These and other unwanted atmospheric parts can be separated based on the temperature at which the steam is condensed into liquid.Due to the proprietary property and boiling point of dielectric fluid, many naturally occurring pollutants can be removed using this method.Removing the fluid that is not easy to condense helps to keep the purity of the dielectric fluid.If the condensation point of fluid is lower than the condensation point of the dielectric fluid under the standard atmospheric pressure by approximately greater than 20 ℃, or if the condensation point of fluid is lower than 10 ℃ under the standard atmospheric pressure, then the fluid will be considered to be not easy to condense.

[0071] During maintenance, startup, and / or shutdown operations, a blanket of inert gas (e.g., nitrogen) may be introduced into the controlled pressure vessel to reduce the amount of dielectric fluid lost when the controlled pressure vessel is opened and / or exposed to atmospheric conditions. Fig.11 As shown, the cooling and steam management system 600 may include an inert gas tank 613 that may supply inert gas to reduce dielectric fluid losses.

[0072] Some disclosed embodiments may include substantially self-contained servers and / or computing systems. In some embodiments, dedicated seals and / or connections may be utilized to reduce the total number of penetrations into the controlled pressure vessel 110. Some embodiments combine power, water, vacuum, and network connections into a single bundle to minimize penetrations into the controlled pressure vessel to reduce the likelihood of leaks when the system is under vacuum.

[0073] Figure 4 An exemplary embodiment of a super structure housing multiple pressure controlled vessels is depicted. In this exemplary embodiment, two pressure controlled vessels 110 are pre-piped, pre-wired, and housed within a modular super structure 210. This allows the embodiment to be prefabricated and delivered as a substantially complete, self-sustaining system. The modular system can be configured to connect to other modular embodiments of the disclosed computing system. In some embodiments, the modular super structure 210 will require only a single power connection and will be pre-wired with appropriate electronics to provide the required voltages to computer components and / or other electronic components.

[0074] Figure 5 An exemplary data center embodiment is depicted showing multiple pressure controlled vessels connected to a central power source. Figure 6 An exemplary data center embodiment is depicted showing multiple controlled pressure vessels connected in series with each other. In these exemplary embodiments, the controlled pressure vessels 110 may or may not be placed within a super structure.

[0075] Fig. 7A -D depicts an exemplary embodiment of a cooling computing system with an internal manipulator, an airlock, and an external manipulator. In this exemplary embodiment, an internal manipulator 230 contained in a pressure controlled container 110 can be used to remove a component 170 and deliver the removed component to an airlock 220. Using the airlock 220, the component 170 can be removed without substantially interfering with or disrupting the pressure, atmosphere, dielectric fluid, and / or other conditions in the pressure controlled container 110. Once the component 170 is removed from the pressure controlled container 110, a replacement component can be introduced into the pressure controlled container 110 using the airlock 220. The replacement component can then be installed by the internal manipulator 230. This process can be significantly simplified using components that can be installed in a "slot-in" manner (e.g., blade servers and chassis).

[0076] A violation of the conditions within the pressure controlled vessel may be detected by a sensor (e.g., a pressure sensor) placed within the pressure controlled vessel. A violation may be indicated by a deviation of at least 10% under conditions outside the standard range of operating conditions. A significant violation of the conditions within the pressure controlled vessel may be indicated by a deviation of at least 30% under conditions outside the standard range of operating conditions.

[0077] In some embodiments, a self-contained diagnostic program can be run that analyzes the performance of components within the controlled pressure vessel 110. If a component 170 is not performing as expected, the robotic arm 230 can be used to automatically remove and / or replace the component. In this way, a self-healing, self-sustaining server and / or computing system can be created. In some embodiments, such a self-healing system can be prefabricated and pre-wired to create a modular unit that can be shipped or delivered to a remote location using conventional methods to provide significantly efficient computing capabilities that require limited setup and / or maintenance.

[0078] In some embodiments, the first steam management goal of cooling and condensing the steam from a gaseous state back to a liquid state is accomplished entirely within the closed system of the vessel by using condensing coils. The process water will be piped through the condensing coils within the vessel. The shape and geometry of the vessel itself will encourage the steam to flow from the bath area to the coil area, and gravity will act to pull the re-condensed liquid back to the bath area.

[0079] In some embodiments, the second vapor management goal of monitoring and maintaining the internal pressure of the container is achieved by using an integrated pressure sensor within the container and using a purge system. In some embodiments, the purge system will be used to remove excess vapor from the container and condense it back into liquid for storage in a liquid storage tank.

[0080] In some embodiments, a third steam management goal of controlling and removing the non-condensable components of steam present during system startup is achieved via the same mechanism as the second goal. A purge system can be used to bring the system under pressure during its initial startup and remove any non-condensable gases from the system.

[0081] In some embodiments, the fourth steam management objective of controlling a blanket of inert gas may be achieved using a dedicated nitrogen blanket feed system. This blanket keeps the coolant below the top of the vessel, allowing for minimization of coolant loss during the phase when the vessel is opened to service components therein. When the operator desires to open the system, dedicated piping from a set of nitrogen storage tanks through a set of dedicated blanket pipes located within the vessel will allow for the addition of an inert blanket. This gas, as well as any other non-condensables, may be removed during the non-condensable removal process that may occur at system startup. The entire steam management process may be managed and monitored through the control system software based on user commands and system status monitoring.

[0082] Ballast block

[0083] In some embodiments of the disclosed system, for example Figure 1 As shown, the pressure controlled container 110 can include a deeper bath portion 142 for accommodating most of the dielectric fluid 140 and a wider shelf area 112 adjacent to the bath. Boards, cards, chips, blades and / or any other computer components 170 are substantially contained in the deeper bath portion 142 of the pressure controlled container 110. The wider shelf area 112 can also accommodate liquid dielectric fluid 140 and / or collect the dielectric fluid 140 that is recondensed into the liquid phase from the vapor phase. In certain embodiments, the depth of the dielectric liquid in the pressure controlled container 110 can be increased by using ballast blocks 160. Ballast blocks 160 can be used to occupy undesirable volumes on the shelves, thereby displacing any dielectric fluid 140 on the shelves 112, and raising the liquid level of the liquid without the need to add additional dielectric fluid 140. In some embodiments, the ballast block 160 includes lift feet 161 that allow fluid to flow beneath the ballast block 160 so that condensed liquid can continue to flow into a deeper bath portion of the controlled pressure vessel without the flow being impeded by the ballast block 160 .

[0084] Ballast block 160 can be made of any material that does not interfere with the operation of the disclosed immersion cooling system. Ballast block can be made of materials including but not limited to metal, rubber, silicone and / or polymer. Preferred materials are substantially insoluble in dielectric fluid. Block must be denser than dielectric fluid, but solid is not required. In a preferred embodiment, block has a handle or cutout, which allows block to be more easily handled and manipulated. Some embodiments of ballast block 160 utilize interlocking top and bottom portions so that blocks can be stacked on top of each other in a safe manner. Interlocking top and bottom reduce the risk of damaging any nearby components if blocks slide or otherwise shift from their predetermined positions. In some embodiments, interlocking top includes a recessed portion that is aligned with the legs and / or lifts on the bottom portion so that the lowest block does not prevent fluid flow, and the block can be safely stacked on the top of the lowest block to occupy a significant volume, thereby allowing the level of the dielectric liquid to be raised without the need to add a large amount of additional dielectric liquid.

[0085] In some embodiments, the ballast block 160 is configured to extend the entire length of the pressure controlled container 110 and / or the shelf 112. In other embodiments, the ballast block 160 can be substantially any size that allows the block to be handled. In such embodiments, multiple modular ballast blocks can be configured to replace as large or as small a volume as desired. In some embodiments, the outer dimensions of a single ballast block are about 2 feet long, or about 3 feet long, or about 4 feet long, or longer, and about 6 inches wide, or about 8 inches wide, or about 12 inches wide, or wider, and about 1 inch high, or about 3 inches high, or about 6 inches high, or about 8 inches high, or higher.

[0086] Super structure

[0087] like Figure 4 As shown, the disclosed computing system is composed of various components, all of which can be directly or indirectly attached to a physical super structure 210. The super structure 210 allows for pre-wiring and pre-plumbing of any required electrical, sensor, control, power, fluid control, pressure control and / or communication systems. This allows for faster and simpler deployment in the field and testing in the factory before delivery to the customer.

[0088] The super structure 210 is typically made of metal components and can be skid mounted or configured to be moved with a forklift, lift, or crane. In some embodiments, the super structure 210 is configured to fit within a standard container to facilitate shipping. The super structure 210 and associated components can be configured to have a total weight of less than about 58,000 lbs and can be divided into smaller subcomponents to facilitate shipping without the need for special equipment. In some embodiments, the weight of the super structure 210 and associated components is less than about 50,000 lbs, or less than about 40,000 lbs, or less than about 30,000 lbs, or less than about 20,000 lbs. In some embodiments, the weight of the super structure 210 and associated components is greater than about 5,000 lbs, or greater than about 10,000 lbs, or greater than about 20,000 lbs, or greater than about 30,000 lbs. Embodiments of the super structure 210 can be of any size and / or shape. Many embodiments are large enough to accommodate multiple controlled pressure vessels 110, server racks 310 and associated liquid immersion cooling equipment, as well as the necessary equipment for managing power delivery and distribution and network connectivity.

[0089] The overall design of Super Structure 210 can be adjusted to accommodate the unique aspects of each deployment, including customizing the type and number of power and process water interconnections to meet the needs of the existing facility.

[0090] The control and management system for all components within the disclosed pressure controlled vessel may be included as part of the disclosed computing system. Preferred embodiments of the disclosed system include all required mechanical systems to maintain and operate a two-phase liquid immersion cooling environment, including required pumps, valves, regulators, steam management systems, pressure management systems, and other associated components.

[0091] The super structure 210 may be an open frame design, or may include side panels and access doors. This allows deployment within an existing structure or outside a field location. The super structure 210 may be modified to include weatherization features, allowing deployment in harsh environments. In some embodiments, the super structure may be a skid / module frame.

[0092] Various systems, features, and / or capabilities may be incorporated into superstructure 210 to support, monitor, and manage other components of the pressure controlled vessel and any environment contained within or associated with the pressure controlled vessel. In some embodiments, such systems may include fire detection and / or suppression capabilities, specialized air conditioning and / or environmental management capabilities, security features such as access control and / or surveillance features, and the like.

[0093] Power System

[0094] Some embodiments of the super structure 210 are designed to accept a variety of electrical input devices and connect them to the existing power distribution system built into the super structure. One of many exemplary embodiments includes a 415V input to a main circuit breaker, which is then distributed to a series of power shelves that convert the 415V AC input to a 12V DC output. In a preferred embodiment, this conversion occurs essentially in one conversion step, thereby reducing the efficiency losses typically associated with such conversions. Traditional computer server locations typically convert incoming industrial power from a high AC voltage (e.g., 415V) to a reduced AC voltage (e.g., 120V). This conversion results in heat energy losses. In general, this can result in approximately 6% energy losses. The 120V power must then be further converted to DC current for use by various computer components. This second conversion results in a second, approximately 6% heat energy loss. By converting approximately 415V industrial power directly to approximately 12V DC, the overall heat energy loss can be reduced.

[0095] Another exemplary embodiment may include connecting a 480V AC input to a power shelf that converts the 480V AC input to a 48V DC output, which is then distributed to a series of intermediate power supplies that convert the 48V DC input to various DC outputs, including, for example, 12V, 5V, 3.5V, 3.3V, etc.

[0096] In some embodiments, there may be a single power supply group, or there may be multiple power supplies operating at different input and output voltages. The exact configuration may be adjusted to meet the needs of the specific device being assembled and depends on the application conditions. The specific design of the power system may be adjusted to meet the needs of the specific environment in which the disclosed computing system is deployed. Customization may include the type, capacity, and interface of the system's power input and output.

[0097] In some embodiments, the rack power distribution system may include a modular power system and / or a group of modular power systems. The specific configuration of the modular power system or multiple systems is not particularly critical, as long as it can deliver the desired amount and type of power to the rack. Therefore, the modular power system can be configured in parallel or in series or a combination thereof to provide one or two or even multiple power distribution paths. The specific path to the rack can be direct or indirect, and can generally depend on the components included, the amount and type of power, and / or the desired configuration. If necessary, the path to the rack can include distributing power to a chassis located within the rack. The distributed power can be delivered at one or more desired voltages, and the voltage can vary according to the configuration and components. In some cases, the desired voltage can include, for example, 12V, 5V, and / or 3.5V. In some embodiments, if a chassis is used, it can use one or more subsystems. Such a subsystem can include any desired subsystem that does not interfere with the amount and type of power to be delivered to the rack. As just one example, a packaged electronic system can be used. Depending on the expectation, this package can accept AC current and convert it into DC current and / or vice versa. For example, particularly useful packaged communication electronic systems may be designed to accept input power at 208, 240, 380, 400, 415, 480 and / or 600 volts AC and convert that power directly to DC power, such as 48V DC.

[0098] The modular power supply system or systems may be powered directly or indirectly in any convenient manner. For example, the modular power supply system may be powered directly via the main power distribution system within the chassis. Depending on the type and number of power and other components, the chassis may use an interface, such as a set of spring-loaded pins or other suitable connector interface, to establish electrical continuity between the power distribution path and the chassis itself. Continuity may then be established between the interface connector and any desired power input interface on the desired server or other computing component located on the chassis. In some embodiments, a packaged power-on module may be utilized within each chassis to convert the voltage to the appropriate level directly in the chassis itself. This may be used for various types of power distribution, but may be particularly useful for, for example, 48V power distribution. Fig.17 An exemplary embodiment of a rack power distribution system 950 is shown. In this exemplary embodiment, a rack 310 may receive an AC input 960 at an AC interface 311 of the rack 310. The power distribution system 950 may generate a DC output 320 and distribute the DC output 320 to one or more chassis 400.

[0099] In some embodiments, ensuring reliable power supply to computer components within the rack is a primary concern. To this end, some embodiments use blade-level power supplies or computer component-level power supplies that can supply a specific input voltage and provide the required output voltage to the blade and / or component-level power supplies. Some embodiments incorporate multiple power supplies to each blade to provide redundancy.

[0100] In some embodiments, one or more switches may require power. An exemplary switch may be a standard data center level switch with appropriate interfaces to connect to the backplane and provide rack level communications to each blade. Some embodiments distribute only a single voltage, and this may be accomplished through a power rail and interface system with connectors to serve as an interface between the power rail and each blade, delivering the voltage directly to the power input rail, or via an intermediate connector placed between the power wires and the rack level voltage distribution system.

[0101] In some embodiments, there may be a power rail that distributes the main voltage along the bottom of the rack. This rail may be powered by one or more main power rectifiers that may be located outside the pressure controlled container and delivered to each rack via a cable or busbar system. Using a higher voltage at this level, such as 48 volts, may reduce the current carrying capacity required by the distribution system and may effectively connect between the distribution rail(s) and the load interface.

[0102] In some embodiments, there will be two main power distribution systems located within the superstructure platform. The first is the primary equipment power system (PEPS) and the second is the auxiliary equipment power system (SEPS). The purpose of the PEPS is to provide electrical services to the components within the container. The system can be a high voltage, high current distribution system that accepts input via a copper conductor or busbar system and delivers the input to the main power source responsible for providing operating current to the chassis, computer components and / or other critical load equipment. Electricity will enter the superstructure at a defined point and terminate at the main service disconnect breaker. Upstream of this point will be electrical services and all power redundancy components used in the system. The input will be at high voltage, such as, for example, 415 or 480 volts AC. The main equipment load will be driven by a power supply or rectifier that is powered by a circuit breaker panel downstream of the main disconnect breaker.

[0103] The purpose of the SEPS is to provide electrical service to all infrastructure support systems and components located within the super structure. Since components required as part of the auxiliary equipment infrastructure may expect a lower input voltage, the SEPS may be powered by a step-down transformer connected upstream of the PEPS main service disconnect breaker via an auxiliary service disconnect.

[0104] This arrangement will allow the superstructure support and infrastructure systems, including all components powered by the SEPS, to be turned on and operated even without primary power being delivered to the rest of the system components. All aspects of the management and control systems, as well as the steam control system, may be able to operate independently of the operation of the PEPS.

[0105] In some embodiments, an uninterruptible power supply (UPS) is included as part of or added to the power distribution system. The incorporation of a UPS allows the disclosed computing system to continue to operate in the event of a temporary interruption of external power.

[0106] The components of the disclosed power distribution system may include, but are not limited to, commercially available components such as, for example, uninterruptible power supplies, DC power systems, AC power systems, and / or power control and monitoring systems. Some such components may include, but are not limited to, Vertiv products such as, for example, Liebert and / or Chloride UPS products, double conversion online UPS, line interactive UPS, backup UPS, lithium-ion battery UPS, and combinations thereof. The UPS products may be single-phase or three-phase. Other exemplary power distribution system components may include, for example, Emerson network power products, NetSure DC power systems, Vertiv, Liebert, Chloride, and / or NetSure power distribution units and related components such as, for example, inverters, rectifiers, transfer switches, and combinations thereof. Commercially available monitoring units, controller units, and / or software associated with these components may also be incorporated into certain disclosed embodiments.

[0107] Controlled pressure vessels and pressure management systems

[0108] An embodiment of the disclosed system includes a pressure controlled vessel designed to house a two-phase liquid immersion cooling system. The pressure controlled vessel 110 contains a bath 142 of a dielectric cooling fluid 140, a condenser 130 with cooling coils 132 to condense the gas phase dielectric fluid into a liquid, and the physical mechanisms and / or equipment required to hold computer components 170 and distribute power from the power system to the devices and components located within the pressure controlled vessel 110.

[0109] During operation, the controlled pressure container 110 may be maintained at a slight vacuum. It should be appreciated that in order to operate a computing system within a controlled pressure container 110 that maintains a negative pressure, various special connections and considerations must be made.

[0110] Some embodiments of the disclosed system use a series of fiber optic Media Transfer Protocol (MTP) interfaces to allow optical fiber to be connected to the controlled pressure vessel 110 in addition to tapping panels and cable trays used to distribute the optical fiber to the rack 310. This arrangement reduces the overall number of penetrations into the controlled pressure vessel 110, thereby reducing the likelihood of leaks in the vessel.

[0111] Some embodiments of the pressure controlled container 110 include sensors to ensure safe operation. These sensors may include, but are not limited to, temperature sensors, fluid level sensors, pressure sensors 180, gas partial pressure sensors, position sensors, electrical sensors, microphones, and / or cameras to ensure and / or automate the operation of the system.

[0112] In an example embodiment, the temperature sensors may include, but are not limited to, sensors for measuring the temperature of the gaseous phase within the pressure controlled vessel 110, sensors for measuring the temperature of the liquid phase within the pressure controlled vessel, sensors for measuring the temperature of water and / or other process fluids, and / or sensors for measuring the temperature of other components including the computer component 170. In some embodiments, thermocouples, thermistors, and / or silicone sensors may be used to measure the temperature of the computer components. In some embodiments, the system may rely on information provided by the components themselves and retrieve or monitor that information using a commonly accepted communication protocol (e.g., an API or other programming interface provided by the device, such as JSON via HTTPT or SNMP) to determine the device temperature.

[0113] Some embodiments may include various life safety features to ensure the safety of the user. These features may include, but are not limited to, automatic electromagnetic locking mechanisms, fail-safe systems, fire and / or smoke detection and / or suppression systems, ventilation systems, and / or backup lighting. In certain embodiments, these features may be combined as part of an integrated platform.

[0114] Certain embodiments include leak detection systems based on automatic vapor detection to ensure that any fluid loss in the pressure controlled vessel is detected quickly. These systems may include a pressure sensor 180 within the pressure controlled vessel 110 that monitors the pressure to ensure that there is no large leak; and / or a gas sensor located on the outside of the pressure controlled vessel that detects the presence of any dielectric vapor that may have leaked out of the pressure controlled vessel.

[0115] The specific design, arrangement and / or layout of the embodiments of the disclosed system can be adjusted based on the conditions in which it is deployed. In some embodiments, the size, materials, internal systems, component mounting and configuration options, interfaces between the pressure controlled vessel 110, computer components 170 and the power system can all be adjusted based on the conditions in which the system is utilized.

[0116] Rack System

[0117] Fig. 8A 400. The rack 310 may be used as a middleware between the electrical and communication systems mounted within the controlled pressure vessel 110 and the computing devices 170 mounted within the rack 310. The computer components 170 may be mounted on the rack 310 to control the spacing, orientation, position and / or configuration of the computer components 170 in the controlled pressure vessel 110. In an example embodiment, each computer component 170 may be mounted in a chassis 400 prior to being mounted on the controlled pressure vessel 110.

[0118] The rack 310 may be any physical structure that may be used to mount the computer components 170, including but not limited to any frame, bracket, support, or other structure. The computer components 170 are considered mounted to the rack 310 when they are directly or indirectly attached to the rack 310 and held in a substantially fixed position. Some embodiments may include using a dedicated mechanical guide plate as a mounting mechanism, a wiring harness attached to a bulkhead assembly, and / or by using an intermediate power supply and backplane receiver 331 to distribute power and signals within the rack.

[0119] The specific design of the rack system 310 can be adjusted based on the conditions of the system deployment. Some embodiments of the rack 310 may include a dedicated switch. In some embodiments, the uplink interface can be connected via a fiber optic infrastructure, and / or the downlink access interface can be connected to the computing device 170 within the rack via a backplane receiver 331 interface or any other suitable means of connecting computing devices.

[0120] In certain embodiments, rack system 310 may include a housing for one or more intermediate power supplies that can distribute the appropriate voltage from the power interface to other equipment mounted within rack 310. Interfaces for interconnecting power from the distribution system to the intermediate power supplies may be incorporated into the design of rack 310 to allow it to be removed and / or replaced with an alternative rack configuration by disconnecting the interfaces between the various rack, power, and communication systems.

[0121] Fig. 8AA top view of a rack 310 is shown. In this example embodiment, the rack 310 includes an AC interface 311 and a data interface 312. The rack 310 also includes a pair of power supplies, a power supply 313 and a redundant power supply 314 (or backup power supply). The rack 310 may also include a rectifier and a controller. The redundant power supplies 314 and / or the rectifier and controller allow the rack 310 to be quickly repaired, or even continue to function if the main power supply stops functioning. The rack 310 may optionally include a converter 315. The rack 310 is configured to receive a plurality of chassis 400 and to maintain the chassis 400 in a substantially fixed position.

[0122] In certain embodiments, whole support 310 can be immersed in dielectric fluid.This can include during operation rectifier, power connection and / or data connection are immersed in dielectric liquid.In order to reduce and / or eliminate the plastic pollution of dielectric fluid, in certain embodiments, plastic insulation and / or cable shielding can be eliminated.In such embodiment, dielectric fluid can be used for insulating cable and / or connection exposed in other ways.

[0123] Figure 8B A perspective view of a rack 310 including a plurality of chassis 400 is shown. The disclosed rack configuration facilitates heat exchangeability of the chassis 400. In this example embodiment, the rack 310 may include a plurality of AC cables 318 connecting an AC interface 311 to a power supply 313 and / or a redundant power supply 314. The power supply 313 and / or the redundant power supply 314 may generate a DC output 320, which may be delivered to a backplane receiver 331 via a DC cable 321. The rack 310 may also include a plurality of data cables 319 connecting a data interface 312 to a backplane receiver 331. The backplane receiver 331 may be used to supply data from a data connection on the bottom of the chassis 400 to a data connection at the top of the rack.

[0124] Figure 8C A side view of the rack 310 is shown. In some embodiments, the rack 310 provides mechanical stability and / or housing for the chassis 400 and its components. In addition, the rack 310 facilitates routing power and data cables from the top of the rack 310 (where they typically enter a container) to the bottom of the rack 310 (where they connect to the chassis 400).

[0125] Chassis and interface systems

[0126] In one example embodiment, the disclosed chassis system 400 is intended to serve as a standardized physical intermediate component between conventional and / or intended built-in computing components 170 and the disclosed rack system 310. In one example embodiment, the purpose of the backplane receiver 331 is to provide a slot-type interface between the chassis 400 and the rack 310, allowing power and signals to be distributed between power supplies in the power system and network switches in the communication system, wherein the various computing components 170 are assembled within the chassis 400.

[0127] In some embodiments, the pressure controlled container of the present disclosure may include at least one rack 310, which may include one or more servers, such as blade servers. Each server may be attached to a chassis 400 (also referred to as a server box or box). Fig. 9A -G shows an example embodiment of a chassis 400 for mounting various components 170. The chassis can facilitate mounting a server on a rack of a pressure controlled container, or removing it from a system. In some embodiments, other electronic components of the pressure controlled container can be mounted in the chassis. For example, computer components or hardware (such as any part of a motherboard, chip, card, GPU, or CPU) can be assembled in the chassis. As another example, components such as a power supply, power interface, or network communication interface can be mounted in the chassis.

[0128] In one example embodiment, a chassis can be used as a common interface between a component (e.g., a server) and a pressure controlled container. The chassis can provide various mounting, power, and connection features that can be customized based on the nature or design of the component. In other words, various aspects of the chassis can be modified based on the design specifications of the component. Thus, the chassis can accommodate nearly any model or type of hardware. For example, the chassis can facilitate the use of specially designed hardware or off-the-shelf hardware.

[0129] Embodiments of chassis 400 may include components designed to allow the adaptation of existing commercial components, the use of custom designed components, and / or the use of specialized chassis for specific applications. Embodiments may include adapter kits for standard motherboards and specialized components. In specific embodiments, such components include a Gigabyte motherboard with an NVidia GPU and / or an Ultra Micro motherboard with an Intel CPU.

[0130] Fig. 9A A chassis 400 for mounting a server on a rack according to an example embodiment is shown. In this example embodiment, the chassis 400 may be a rectangular box including a rear wall 410 and two side walls 420. The rear wall 410 may include a plurality of holes 411 to facilitate fluid circulation within the chassis 400. The chassis 400 may include guide rails 421 on each side wall 420.

[0131] Fig. 9B 4 shows various components inside the chassis 400 according to an example embodiment. In this example embodiment, the rear wall 410 is removed. Thus, Fig. 9B A server 430 is shown that includes a power module 431, a GPU module 432, a CPU module 433, and an interface card 434. In an example embodiment, the components inside the chassis 400 may include components used in a blade server, such as the CPU module 433 and the GPU module 432. In addition, the components inside the chassis 400 may include other components that are not traditionally included in a server, such as the power module 431 or the interface card 434. Because the chassis 400 does not require traditional air cooling equipment, the chassis 400 does not include any fans or heat sinks in the chassis. Therefore, the chassis has a very thin profile relative to the computing power of the chassis.

[0132] Fig. 9C Schematic diagram of components within the chassis is shown. In this exemplary embodiment, a server motherboard 445, a plurality of power modules 431, and an interface card 434 are mounted on the chassis 400. Storage devices and / or other peripheral components may also be mounted to the chassis 400 along with the backplane interface 330 and / or the power modules and communication system modules.

[0133] In one example, a mounting interface can be added to or removed from the chassis so that a piece of hardware is fixed to the chassis. On the inner surface of the chassis 400, a device that allows components (e.g., motherboards, GPUs, CPUs, interface cards, and other related components) to be mounted to the chassis can be formed. These devices are mounting interfaces. The specific arrangement of the chassis system 400 can depend on the equipment and / or components to be attached to the chassis 400 and / or the rack. Some embodiments of the chassis 400 can have interchangeable mounting plates that can be used for device attachment. A set of standard attachment plates can be used for common or commonly used components.

[0134] The style and form factor of the power and network interface modules in the chassis system 400 can be adjusted based on the instructions and requirements of certain components and / or user-specified devices. In one example, the power subsystem of the chassis can be modified to meet the needs of specific components. In another example, the size of the chassis can be designed to adapt to a piece of hardware of any size. In yet another example, the chassis can provide different network options according to the network connection card assembled in the chassis. Because of these features and other features of the chassis, the chassis can adapt to various components. Therefore, the assembly or removal of these components of the pressure-controlled container can become simplified, thereby automated. For example, the chassis can include a blade server, and the robot can easily assemble the chassis to the rack of the pressure-controlled container or remove the chassis from the rack of the pressure-controlled container. Therefore, the robot can remove and replace the blade server without any manual interaction, thereby minimizing human exposure to dielectric fluids.

[0135] In one example embodiment, the chassis may include a microcontroller that may communicate with a management system of the pressure controlled container. The microcontroller may receive sensor data from various sensors placed in or outside the chassis. For example, the chassis may include a sensor for detecting whether the chassis is properly placed in a rack. If a server can be connected to the rack, the server may be properly placed in the rack. The sensor may determine whether the chassis is properly placed in the rack. Thus, the sensor may transmit data to the microcontroller, and by using the data, the microcontroller may provide a signal to the management system indicating whether the chassis is properly placed in the rack.

[0136] In one embodiment, the microcontroller may be coupled to a switch that may turn on or off a component mounted in the chassis. The microcontroller may receive a power on or off signal from a management system, and in response to receiving the signal, the microcontroller may transmit a signal to the switch to turn on or off a component, such as a server. In an example embodiment, the microcontroller may receive operational data from a server, and the microcontroller may forward the data to the management system. The operational data is a key performance indicator of a server and may be indicative of its performance. The operational data may include computing speed, computing degradation, power consumption, temperature of a circuit, and bandwidth of a system.

[0137] In one example embodiment, the microcontroller can monitor, manage and control the electrical and communication facilities of the blade server. For example, monitor indicators of current (i.e., amperage) and voltage to ensure that the system can protect itself, for example, not providing overcurrent or undercurrent.

[0138] In one example embodiment, the chassis may include a structure that enables a robot to grasp and remove the chassis. For example, the chassis may be in the shape of a rectangular box having a front, a rear, and side walls. The chassis may also include a top wall and a bottom wall. The top wall of the chassis may include a plate that may be coupled to a robotic arm. Using the plate, the robotic arm may grasp the plate for unloading and other handling operations.

[0139] In one example embodiment, the chassis may include mechanical guide rails and locating pins to ensure proper alignment and insertion of the chassis in the rack. The mechanical guide rails may be placed on the side walls of the chassis.

[0140] In one example embodiment, the chassis can include various features to promote fluid flow. For example, the chassis can be in the shape of a rectangular box having a front, a rear, and side walls. The chassis can also include a top wall and a bottom wall. In this example, at least one wall of the chassis can include fluid flow openings extending through the wall. For example, the rear wall can include a plurality of openings that can facilitate fluid flow into and out of the chassis when the chassis is immersed in a bath of liquid.

[0141] In one example embodiment, the chassis may include perforations to ensure that all fluid within the chassis drains when the chassis is removed from the liquid bath. For example, a rack may be located in a liquid bath to cool computer components held by the rack. To remove a server, a robot may grab a plate of the chassis and lift the chassis from the rack (thereby removing the chassis from the liquid bath). When the chassis is removed from the liquid bath, a certain amount of fluid may remain within the chassis. The chassis may include notches or drains at the bottom wall of the chassis to ensure that the fluid can escape even if the pressure controlled container is not completely horizontal. The notches or drains may be located at the corners of the bottom wall.

[0142] In an example embodiment, the chassis may include a power interface and / or a communication interface. The interface may electrically connect components installed within the chassis to a rack and / or a pressure controlled container. The power interface and / or the communication interface may be placed at a backplane. For example, a server installed within the chassis may be connected to an interface of the chassis via various wires and cables. When the chassis is placed within the rack, the interface may be electrically connected to another interface connected to the rack (i.e., a backplane receiver) and / or a pressure controlled container. The electrical connection between the two interfaces (e.g., a backplane and a backplane receiver) may provide power to the server and connect the server to a communication network within or outside the pressure controlled container. The connection between the two interfaces may be automatically completed during the mechanical insertion of the chassis into the rack. Similarly, removing the chassis from the rack may disconnect the server from the rack and / or the pressure controlled container.

[0143] In some embodiments, providing standardized interconnections through the backplane interface 330 and the communication system interface can minimize the possibility of incorrect connection of the data interface and reduce the need for connection troubleshooting.

[0144] In some embodiments, the chassis 400 will include a set of standard power and network interfaces. The network interface can be in the form of a Cat6A or Cat7 compatible RJ45 interface for connecting to a 1G or 10G Ethernet interface on the device motherboard. In such an embodiment, the power interface can include a set of standard Molex type connectors for connecting to standard motherboards and / or peripheral components.

[0145] In one example embodiment, the pressure controlled vessel may include an internal database for storing information about components assembled within the system. The internal database may be a repository of components assembled on the pressure controlled vessel. For example, the internal database may store the brand and model of each server and power supply assembled within the system. When components of the system are replaced or replaced, such as by a robot, the management system may track the changes and update the information stored in the internal database. The pressure controlled vessel may also be connected to an external database via a network.

[0146] In one example embodiment, each chassis may be associated with a unique serial number, for example, displayed as a bar code on the chassis. When a component is placed within the chassis, the specifications of the component (or the brand and model of the component) may be stored in an external database associated with the unique serial number. Subsequently, when the chassis is assembled in a pressure-controlled container, the pressure-controlled container may find the component by searching the external database for the unique serial number. For example, a robotic arm may scan the bar code on the chassis, and the management system may use the bar code to search the external database. The management system may update an internal database using information obtained from the external database. Similarly, when the chassis is removed from the pressure-controlled container, the robotic arm may scan the bar code associated with the chassis, and the management system may update the internal database to indicate that the component installed in the chassis is no longer assembled in the system.

[0147] In one example embodiment, the chassis may include an RFID tag. The robotic arm of the pressure controlled container may include a scanner capable of transmitting a radio frequency to detect the RFID tag. When the robotic arm handles the chassis, the robotic arm may scan the RFID tag and provide a unique serial number to the management system to update an internal database.

[0148] In one example embodiment, the chassis may include an identification plate that may contain an asset identification number specified by a user. The asset identification number may be stored in association with a component installed in the chassis. In some embodiments, the identification plate may be a chip configured to store the asset identification number.

[0149] In one example embodiment, the chassis may include a pump to enhance fluid flow within the chassis. To maximize heat exchange between components within the chassis and the liquid bath, the chassis may include a pump that circulates fluid within the chassis and around the components. The pump may draw fluid from various conduits distributed around the chassis and drive the fluid outside the chassis, or vice versa.

[0150] In one example embodiment, the chassis may include various conduits around the chassis for drying the chassis and components mounted therein. When the chassis is pulled out of the liquid bath, a certain amount of liquid may remain in the chassis or in the components therein. The chassis may include various conduits that can direct airflow within the chassis or around the components to facilitate drying of the chassis and components. In one example embodiment, a pressure-controlled container can expose the chassis to airflow before delivering the chassis to a user. For example, the chassis may include an input conduit for receiving an airflow, and the pressure-controlled container provides airflow through the input conduit.

[0151] Fig.9DA bottom wall 415 of the chassis 400 is shown according to an example embodiment. In this example embodiment, the bottom wall 415 may include a power interface 416 and a communication interface 417. Fig.9D Guide rails 421 on the side walls 420 of the chassis 400 are also shown.

[0152] Fig.9E A top wall 425 of a chassis 400 is shown according to an example embodiment. In this example embodiment, the top wall 425 may include a plate 426 and a pair of handles 427. The plate 426 may be used by a robot arm to pick up the chassis 400.

[0153] Fig.9F A side wall 420 of a chassis 400 according to an example embodiment is shown. In this example embodiment, the side wall 420 may include a guide rail 421. Fig.9F Also shown are rear wall 410 , handle 427 , and power port 416 .

[0154] Figure 9G An exploded view of a bottom drain opening 450 of a chassis 400 according to an example embodiment is shown. In this example embodiment, the bottom drain opening 450 may be placed on a corner of the bottom wall 415, the side wall 420, and the rear wall 410.

[0155] Fig. 10A -F shows an example embodiment of a pressure controlled container 500. In particular, Fig. 10A An example embodiment of a container 500 is shown, for example, a 600KW skid device. This example embodiment includes a modular skid. The container 500 may include a plurality of fork lift tubes 514 that facilitate the movement and transfer of the container 500 to a desired location. The container 500 may receive power and communication inputs 511, as well as process water from a process water pipe 512 that minimally penetrates the container itself. These connections may be positioned on the top of the container to facilitate the closed packaging of the modular container in the data center. In some embodiments, the connection may be placed on the front and / or side of the container to accommodate the vertical stacking of multiple modular containers in the data center. In some embodiments, the container may include vertical spacers to facilitate the vertical stacking of containers on top of each other. The vertical space may create additional space for connections, air flow, and / or insulation between containers. By stacking containers vertically, extraordinary power density may be achieved on a square foot basis. In some embodiments, the container 500 may include a power and communication box configured to receive inputs 511 and distribute power and network connections throughout the container 500. The container 500 may include a sealing lid 515 , which may facilitate adding and / or removing components from the container 500 .

[0156] Fig. 10BAnother view of the container 500 is shown. In some embodiments, an inventory of replacement parts can be stored within the container 500 so that the parts can be replaced using a robotic system within the container without opening the container. The robotic system can be operated using a gantry motor 516. In such an embodiment, when a part breaks or needs repair, the replacement part is assembled into the system, and the broken or otherwise removed parts can be stored in the cassette until the cassette is full. At this point, the cassette containing the removed parts can be removed from the container, and a fresh cassette with new replacement parts can be inserted into the container for future use. In some embodiments, the disclosed container is approximately 15 feet long, approximately 7 feet wide, and approximately 10 feet high. In some embodiments, the disclosed system can provide 600KW of computing power to be implemented in approximately 150 square feet.

[0157] In some embodiments, the container 500 may also include one or more bellows tanks 517. The bellows tanks 517 may be used to regulate the pressure in the container. When the disclosed computing and / or cooling system is initially activated, the expanded dielectric fluid may be directed to the bellows tanks so that it is not lost to the environment and / or to avoid pressure accumulation in the container. In some embodiments, the bellows tanks 517 may be large enough to accommodate approximately twice the amount of liquid dielectric fluid in the container.

[0158] Fig. 10C A cross-sectional view of a container 500 is shown. The lower portion of the container 500 can accommodate racks 310 and / or chassis 400, which house computing components. Above the racks are condenser coils 132, which cool and condense any dielectric vapors. Power can be distributed within the container using power busbars 518. This allows power to be distributed to individual computing components in a hot-swappable manner. The power busbars 518 allow the container to receive external power using only one or a small number of penetrations through the container. This design simplifies the assembly and operation of the container system. In some embodiments, each power busbar can provide 600 amps of current to the power supply of five racks. In such an embodiment, there can be two sets of busbars, one set on each side of the container. In some embodiments, the busbars do not include plastic insulators. Plastics may be considered a contaminant of certain dielectric fluids and are generally avoided in some embodiments.

[0159] In some embodiments, the container 500 may include a desiccant 519. In some embodiments, dielectric vapor may be removed from the head space of the container 500 and condensed in a manner that allows any non-condensable components to be removed from the dielectric fluid. Under the same conditions as many dielectric fluids, water will not condense. Therefore, the system may be used to remove water contaminants from the dielectric fluid.

[0160] In some embodiments, container 500 may include fluid filter 520, fluid conduit 521 and fluid pump 522. In some embodiments, dielectric fluid may be added to container in a manner that liquid dielectric fluid overflows rack 310 and enters into liquid collection area 523. Then, fluid filter 520 may be used to filter fluid, and fluid pump 522 and fluid conduit 521 may be used to pump fluid to the far side of container. The system circulates newly filtered dielectric fluid through container, so dielectric fluid may be reused for cooling computing components.

[0161] Fig. 10D A cross-sectional view of the container 500 is shown. In this embodiment, the level of the liquid dielectric fluid can be maintained at a fluid level 524 that is higher than the height of the rack 310 and / or the computing components therein. Therefore, the rack 310 and / or the computing components can be immersed in the dielectric fluid. Above the fluid level 524 can be saturated dielectric vapor, for example, up to an intermediate level 525. In some embodiments, the saturated dielectric vapor is maintained to an intermediate level 525, which can be approximately half the height of the condenser coil 132. Above the saturated vapor is a head space, which in some embodiments can contain a lower density dielectric vapor.

[0162] exist Fig. 10D In an exemplary embodiment of the present invention, the cooling coil 132 is located above the shelf area. Therefore, when the robot 526 places or removes the chassis 400, the cooling coil 132 is out of the way. This arrangement of the cooling coil 132 can simplify the placement and removal of the chassis 400, which can provide significant benefits in the autonomous operation of the container.

[0163] Communication System

[0164] Embodiments of the disclosed communication system are designed to provide standardized Layer 1 to Layer 3 connectivity and management interfaces for devices within or associated with the disclosed superstructure 210, controlled pressure vessel 110, and / or computing system.

[0165] In some embodiments, a series of MTP interfaces provide the ability to connect multiple high-density multimode optical fibers into the controlled pressure vessel 110. Once contained in the controlled pressure vessel 110, the fiber connections can be segmented into separate switch-level connections using a set of dedicated drop cables, drop interfaces, patch panels, and / or distribution patch panels to the rack 310.

[0166] Some embodiments of the disclosed system may include a dedicated fiber optic patch panel interface port at each rack 310 to allow connection via a short patch panel to a switch system mounted therein. In other embodiments, there may be a dedicated patch panel or set of patch panels extending from each switch system to the MTP distribution interface.

[0167] In some embodiments, the interface between the switch system and chassis 400 may be via backplane interface 330 and / or via some other mechanism, which may or may not include the use of backplane connectors. In some embodiments, there may be no intermediate rack-level switch system. These embodiments may use a set of centralized switches within the pressure controlled vessel 110 to connect to the various computing devices located therein.

[0168] A typical interface between the switch system and chassis 400 may be implemented using patch panels attached to rack 310 and routed to backplane system 330 using patch cables that connect ports on the patch panels with appropriate ports on the switch system.

[0169] In some embodiments, there will be a small (6U) rack rail area containing patch panels that interconnect the communication system cabinets with the MTP interfaces on each controlled pressure vessel 110, as well as a centralized communication system distribution switch that is used to interconnect the switch systems to each other and / or to the outside world. In such an embodiment, the end user or customer may choose to rig up their own routing equipment within this space and provide it with external connections to serve as connections between the disclosed computing systems and the outside world, or to run fiber optic connections between the controlled pressure vessel 110 or super structure 210 and an existing network environment.

[0170] The access, communication and / or networking components used in embodiments of the communication system environment may be standard equipment or may be user specified. The rack 310 and backplane interface 330 system may include the ability to replace the switch system located within each rack 310 by removing the existing switches, replacing the existing switches with any standard switches (e.g., 1U switches), and rewiring the desired interfaces to the backplane network interface panel.

[0171] In certain embodiments, products may be utilized that are designed to interface directly with the backplane system 330. Such products may utilize the chassis 400 patch panel system and / or a direct electrical interface specifically designed to interconnect switch ports via a purpose built network interconnect interface, via a commercially used protocol, or via a specification for designing a network level interconnect interface.

[0172] In some embodiments, the connection between each blade or chassis and the switch may include multiple interfaces. One interface may be a standard switch port, which may be a standard port available on a commercial switch. A common interface may be 1GBASE-T or 10GBASE-T, which utilizes a Cat6 or Cat7 twisted pair copper connection between the switch and the host device. Another interface may be an intermediate device from the switch to the backplane, which may be composed of a patch panel and a set of hardwired connections, the patch panel having a standard patch cable extending from a standard switch port to the front side of the patch panel, and the hardwired connection from the back side of the patch panel to the signal interface of the signal backplane. Alternatively, this may be composed of a dedicated cable and / or a standard RJ45 interface extending from the switch port to the board to establish a connection between the standard switch port and the backplane. Another interface may be an interface system signal backplane, which distributes signal paths from a standard switch port along a printed circuit board (PCB). One or more signal paths may terminate at a connector on the PCB, to which the signal backplane interface will be connected. Yet another interface may be a chassis signal backplane interface. This may be a connector located on the chassis itself that mates with a connector on the interface system signal backplane. It serves as an interface between the interface system signal backplane and the chassis itself. Another interface may be the chassis network interface. This may be a standard patch interface that allows a patch cable to be connected from the chassis network interface to an RJ45 connector on a server that is attached to the chassis.

[0173] Robotic system

[0174]

[0046] In some embodiments of the disclosed system, one potential approach addresses the need for heat exchangeability of components within the controlled pressure vessel 110. The need for the ability to remotely remove and replace a failed component 170 may be addressed through robotics.

[0175] Specific embodiments of the combination of the disclosed systems may include an internal robotic arm 230 and / or an external robotic arm 240. Some embodiments, such as those used in cryptocurrency applications and / or certain high performance computing environments, may not require the ability to hot swap components. In other hyperscale GPU and CPU environments, this may be an essential requirement. Embodiments of the disclosed robotic system allow chassis and / or other computer components to be replaced without disrupting any other components. In some embodiments, failed cards and / or components can be automatically and / or planned replaced and / or stored. This allows embodiments of the disclosed system to operate in the short and medium term, fully remotely and autonomously.

[0176] The internal robot 230 mechanism is located within the pressure controlled container 110 environment. Fig. 7A-D, in an exemplary embodiment, when a card or component is not functioning properly, a removal sequence can be initiated. When the removal sequence is initiated, the internal arm 230 will remove the appropriate computer component 170 and / or associated chassis 400 from the rack 310, move it to the airlock 220 located within the pressure controlled container 110, and send a signal that the removal sequence is complete. Once the sequence is complete, the internal airlock door 222 will close, the airlock pressure will balance with the pressure of the external atmosphere, and the external airlock door 224 will open. Once the external door 224 is open, the external robotic arm 240 will remove the chassis 400 from the airlock 220 and place it in an empty storage slot.

[0177] In some embodiments, the airlock 220 will be purged with nitrogen and / or another inert gas and / or non-condensable gas before the airlock 220 is opened to the external environment. In such embodiments, this has the effect of reducing or eliminating dielectric vapor loss when the airlock is opened and closed. In certain embodiments, the airlock will be equipped with a one-way valve on the inside, outside, or both. In embodiments where both the inside and outside of the airlock have one-way valves, purging the airlock will prevent cross contamination of the external environment into the internal atmosphere of the pressure controlled vessel 110, and also prevent loss of dielectric vapor.

[0178] When a card or component replacement sequence is initiated, the external robot 240 will remove the replacement component and / or chassis 400 from the storage slot and place the component into the airlock 220. Once completed, the external airlock door 224 will close, the airlock pressure will equalize with the pressure inside the pressure controlled container 110, and the internal door 222 will open. Once the internal door 222 is open, the internal robot 230 will remove the chassis 400 from the airlock 220 and insert it into the appropriate rack 310.

[0179] When coupled with a remotely accessible management system, the internal and external manipulators 230 and 240 allow for remote operation and management of a data center environment. This can reduce the need for human operators, thereby maintaining availability and reducing costs and / or downtime. In some embodiments, the external manipulator 240 is mounted on a movable base, allowing a single external manipulator system to service multiple embodiments of the disclosed computing system.

[0180] When integrated with a custom-developed workflow management system and virtualization technology, the disclosed robotic system allows for the development of a fully autonomous, self-healing data center solution that can provide maximum system reliability.

[0181] In some embodiments, an asset tag with a unique human and / or machine readable serial number and / or production batch code may be included on each computer component and / or chassis. In these embodiments, the asset tag may be a unique serial number. The tag may include a printed bar code or QR code and allow automatic partial identification by an embodiment of the disclosed robotic system. The tag code may also be used in conjunction with a management software system to provide detailed component information about warehouse management and automated systems. The tag and any associated adhesive or other components are preferably made of materials compatible with dielectric fluids. The tag is preferably located on the chassis, in a readable place when the chassis is inserted into a rack. In some embodiments, secondary tags or additional tags may be located in other areas of the chassis to help identify components and / or inventory management.

[0182] Embodiments of the disclosed robotic system allow any individual chassis to be temporarily removed and replaced, a process known as "re-seating". This is useful during troubleshooting when it is determined that a hard power cycle of a component is required. Re-seating accomplishes this by disconnecting all power, waiting a few moments, and reconnecting.

[0183] Some embodiments allow individual cards and / or chassis to be removed from a pressure controlled container through an airlock. In some embodiments, a robotic system will remove the chassis from a slot in the rack, move it to the airlock, and signal the completion of the task to allow the airlock to be opened and the card and / or chassis to be removed. Some embodiments allow replacement parts and / or chassis to be placed into a specific rack slot through the same airlock used for removal. In some embodiments, a robotic system will remove the chassis from the airlock, place it into the appropriate rack slot, and signal the completion of the task.

[0184] Robots in internal systems

[0185] Embodiments of the disclosed system may include an "internal robot" robotic system. In such an embodiment, the pressure controlled container may be expanded to accommodate a robotic arm operating within the container. The container may also be arranged to accommodate movement or transfer of computer components and / or chassis above a rack containing the operating computer components. It should be understood that the pressure controlled container may also be referred to as a tank, a cabin, and / or a vacuum chamber. Alternatively, it is understood that certain components of the pressure controlled container may be referred to as tanks or cabins.

[0186] Fig. 10EAn embodiment of the disclosed system is depicted, the system having a gantry robot 526 configured to remove, replace, and / or assemble computing components (e.g., chassis 400 of rack 310). In some embodiments, the gantry robot 526 may also be configured to remove, replace, and / or assemble DC rectifiers and / or other components of the power distribution system. It should be understood that some embodiments of the disclosed computer components and power distribution components may be designed to be hot-swappable and may include handles or other features that facilitate handling by the gantry robot 526. In some embodiments, the gantry robot 526 is arranged to travel in the x and y directions and may be lowered in the z direction to remove and / or assemble replacement components. In some embodiments, the gantry robot 526 includes a gripping tool to grip the chassis 400 and / or power supply, for example, the gripping tool may grip the board 426.

[0187] Fig. 10E A top cross-sectional view of an example embodiment of the disclosed storage tank is shown. In some embodiments, an array of racks 310 can be populated with chassis 400 and / or computing boards. In some embodiments, each chassis 400 can utilize approximately 6KW of power, and each rack 310 can contain 10 chassis. Therefore, in an embodiment containing 10 such racks 310, the container can utilize approximately 600 kilowatts of power for computing purposes. In some embodiments, additional racks 310 and / or cases 527 of chassis 400 and DC power rectifiers can be stored in the container 500 to serve as replacement parts and / or provide space to store parts that have been removed from the container 500.

[0188] Robots on external systems

[0189] Fig. 12A -E shows another embodiment of a container. Specifically, Fig. 12A An embodiment of a container 700 is depicted in which a gantry robot 526 is external to a tank 710 containing a chassis 400 and / or computing components. In this embodiment, the tank 710 may be smaller, but may need to be opened more frequently to allow the external gantry robot 526 to access the chassis 400 and / or power supply inside the tank 710. Additionally, replacement equipment may be stored and / or contained within a modular enclosure, such as a reservoir 716 external to the tank 710. In some embodiments, the tank 710 may have multiple doors 711 to limit exposure to the interior of the tank 710 when a single door 711 is opened to remove, assemble, and / or replace components or the chassis 400. In such an embodiment, replacement components may be stored external to the tank 710 to avoid unnecessary opening of the tank.

[0190] In addition, the vessel 700 may include one or more transformers 712, electrical distribution panels 713, process water pipes 512, and electrical ducts 714. The vessel 700 may also include a programmable logic controller (PLC) cabinet 715 that monitors and controls the status of various devices within the vessel 700. The transformer 712, electrical distribution panel 713, process water pipes 512, electrical ducts 714, and PLC cabinet 715 may be located outside the tank 710.

[0191] Fig. 12B A cross-sectional view of a container 700 is shown wherein an external doorframe robot 526 may enter a storage tank 710. In this example embodiment, the condenser coils 132, the rack 310, and the bellows 717 are located within the storage tank 710. Fig. 12C A side view of a container 700 is shown, wherein a tank 710 has an external door frame robot and a plurality of doors 711. In this example embodiment, the tank 710 includes a fluid pump for removing fluid from the liquid collection area and sending the fluid to the fluid filter 520 through the fluid conduit 521. The container 700 also includes a casing 718 for storing replacement equipment. In this example embodiment, the casing 718 is located outside the tank 710. In some embodiments, spacers and / or ballast blocks 160 can be used to reduce the total volume of the liquid dielectric fluid in the tank 710.

[0192] Fig.12D A rack 310 is shown according to an example embodiment. In some embodiments, a redundant power supply 314 can be positioned on an opposite side of the rack 310, rather than adjacent to the primary power supply 313. Additionally, power and / or data cables 318 and 319 can be routed in alternative configurations to accommodate the specific requirements of a particular deployment. In this example embodiment, a backplane receiver 331 is located below the rack 310.

[0193] Fig.12E An example hinged door 711 is shown that can be used in some alternative embodiments of the disclosed tank 710. In some embodiments, a sliding door can be utilized instead of a hinged door to reduce or avoid induced currents in the dielectric vapor. Slowly sliding the door open may disturb the dielectric vapor less than opening a hinged door and causing mixed currents.

[0194] Management System

[0195] The management system is a web-based interface between users of the disclosed computing system and the computing system itself. Embodiments of the management system provide an operational view of the computing system and allow monitoring and management of various components, including monitoring and management of the pressure controlled vessel 110, the robotic system, the communication system, the power system, and / or all other systems and components. In an example embodiment, the management system may be implemented in Fig. 12AIn another example embodiment, the management system may be implemented in a PLC cabinet 715. Fig. 10A The power and communication box 513. In each embodiment, the power management system can be implemented as a software program on a control device or other suitable device (e.g., a computer).

[0196] In certain embodiments, a set of simple network management protocol accessible data points may be made available to users of the management system to allow monitoring of key operating parameters via third party monitoring systems. A full operating log may be maintained and graphs may be provided for users to review operating condition data.

[0197] Regular maintenance of system components can be scheduled and maintained via the management system. Users can be provided with regular reminders of routine maintenance and be informed in the interface when regular maintenance is performed. This data can all be maintained as part of the operation log information for historical operation review.

[0198] In some embodiments, operational functions may also be exposed via an API interface to allow remote programmatic monitoring and management of the computing system and related components. A full set of operational monitoring and alarm functions may be included to notify operators when any problems occur.

[0199] A centralized server version or a hosted cloud-based management version of the management system can be used by customers with multiple pressure controlled vessel computing systems. This provides operators with a single program and user-accessible interface for managing a fleet of pressure controlled vessel computing systems.

[0200] In some embodiments, software-based interface modules allow interoperation with computing platforms and third-party management utilities such as Microsoft System Center and VMWare VCenter. The user and API interfaces provided by the management system can allow full interoperability with the disclosed robotic system, allowing full remote and programmatic autonomous operation and supervision of the disclosed computing platform.

[0201] In some embodiments, the control system allows for regulation and control of operations, including temperature, pressure, flow rate, and / or power management. In some embodiments, a user authentication system allows multiple unique users to authenticate to the system. Some embodiments include a role-based and / or element-based permission system. In such embodiments, an administrator would be able to configure multiple roles, with which users can be associated and / or apply specific permissions to individual users outside of their role assignments.

[0202] Some embodiments incorporate video management to provide a user with the ability to record and retrieve video feeds from cameras that may be located within the container and / or super structure. In some embodiments, the camera may acquire visual data that may be analyzed by the processor. In such embodiments, the processor may utilize computer vision techniques to control the operation of the container, robot, and / or super structure system in response to the acquired visual data.

[0203] In some embodiments, the control system and software may be configured to generate reports regarding the operation and status of the entire system and / or individual subsystems and / or components of the disclosed computing platform.

[0204] Exemplary Combined System Embodiments

[0205] It should be understood that the disclosed systems can be used alone or in combination. There are multiple embodiments of the combined computing system, which can be customized for various use cases.

[0206] One example embodiment is the Crypto Series. This is an ultra-high density implementation of the disclosed technology that utilizes purpose-built computing hardware, a rack 310 with a bootstrap and wiring harness designed for that hardware, a modified implementation of the communications system 360 architecture, and a 1 MW controlled pressure vessel 110 and power distribution system. Typical users of this embodiment are those who wish to perform cryptocurrency mining or other ultra-high power density processes using custom computing components, or computing component manufacturers who wish to develop a full range two-phase liquid immersion cooling system into which they will incorporate their own hardware.

[0207] Another example embodiment is a GPU family. This is an implementation of high-density GPU supercomputing of the disclosed technology. This implementation will utilize customized chassis 400, rack 310, and backplane interface 330 technologies designed to combine Gigabyte's motherboard and NVidia's GPU using NVidia NVLink technology to facilitate ultra-high-speed GPU-to-GPU communication. Typical users of this technology include general-purpose parallel processing applications that can take advantage of GPU-based computing and memory capabilities, including graphics rendering, particle simulation, and general research activities.

[0208] Another example embodiment is the CPU family. This is an implementation of high density CPU computing of the disclosed technology. This implementation will utilize high end Supermicro based motherboards, Intel Xeon CPUs, high speed network interfaces, high speed memory and solid state storage for local storage. Typical users of this technology include data centers, enterprises and cloud / VPS hosting providers and service providers who use high performance computing for their own internal applications or for applications provided to third party customers and other organizations.

[0209] Yet another example embodiment includes the Edge Series. This is a scaled-down implementation of the disclosed computing system that is specifically designed for remote / on-site deployment, either within or alongside traditional commercial and data center environments. This embodiment focuses on a secure, weather-resistant environment with full remote monitoring and management capabilities. The target users of this technology will be operators of on-site deployments and distributed technologies, such as network operators and other organizations with distributed on-site infrastructure, as well as operators of existing facilities that want to enhance their computing capabilities with minimal modifications to existing facilities or structures. The system can make various modifications to the external structure to simplify the connection of utilities, including the system's electrical, water, and network connections.

[0210] Self-sustaining embodiment

[0211] Some disclosed embodiments do not rely on an external water source. Such embodiments may include a closed-loop chiller for cooling water or another fluid that may be circulated through a condenser as described above. Using a closed-loop chiller rather than an external cooling water source allows the embodiments to be substantially self-sustaining.

[0212] Fig.13 An exemplary self-contained container 750 is shown. Fig.13 The example embodiment utilizes a skid mounted closed loop chiller 719 to cool water or another liquid to be used in a condenser within a cabin or submerged tank 710. By using a closed loop chiller, the need for an external cooling water source can be eliminated, resulting in a self-sustaining data center solution that requires only external power and network connectivity to fully operate. The container 750 may also include a bellows 717, a door 711, a gantry robot 526, an electrical distribution panel 713, a PLC cabinet 715, and a case 718.

[0213] In some embodiments, the closed-loop cooler 719 can be a skid-mounted closed-loop cooler that is enclosed in the outer shell of a modular pressure-controlled container. In such an embodiment, heat is transferred from the computer components to the dielectric liquid in the tank 710. As described herein, the process converts the dielectric liquid into a dielectric vapor. The dielectric vapor rises in the tank 710 and is cooled by the condenser, thereby converting the dielectric vapor back into a dielectric liquid. The heat transferred from the dielectric vapor to the condenser is then transferred from the condenser to the refrigerant or condensing fluid in the condenser, and then transferred to the closed-loop cooler 719. In some embodiments, the cooler 719 uses vapor compression, a compressor, an evaporator, a heat exchanger, or other closed-loop methods for cooling the refrigerant or condensing fluid to remove heat from the refrigerant or condensing fluid. The heat from the refrigerant or condensing fluid is ultimately dispersed via air cooling. In some embodiments, this results in a self-sustaining, modular, air-cooled two-phase liquid immersion computing system. The air cooling of some self-sustaining embodiments is surprising because the field of immersion cooling is generally opposed to air cooling, especially air cooling of self-sustaining equipment.

[0214] Some disclosed embodiments may be provided in a form factor that saves space footprint. One example embodiment includes a single rack containing ten blades or servers immersed in a dielectric liquid as described above. In some embodiments, each server may draw approximately 6 kW of power. Thus, some example embodiments provide approximately 60 kW of computer power in a small footprint.

[0215] Fig.13 The illustrated example embodiment is contained in a footprint of about 4 feet 2 inches deep, about 8 feet 8.5 inches wide and about 8 feet 8 inches high. This example embodiment includes a computer power of about 60kW and other operating components and systems, and it is contained in an area of ​​about 36.3 square feet. It should be understood that the operating components of the container may include but are not limited to a storage tank or cabin that holds dielectric fluid, a condenser, a power supply and a data connection for computer components. This container may also include a sensor, a control device, an electric power cabinet, a bellows 717, a vacuum system, a fluid filter, a purification system and / or other components. Some self-sustaining embodiments may include an outer shell. In certain embodiments, the outer shell may enclose a container, provide structural support, be skid-mounted, ventilated, weather-resistant and / or waterproof and / or decorative. In certain embodiments, the outer shell of the independent container may include a radiator coil, a fan grille, a heat transfer component and / or an air cooling component for the convenience of using a closed-loop cooler.

[0216] In some embodiments, the self-sustaining computing system provides computing power of at least about 1.5 kW per square foot, at least about 1.6 kW per square foot, at least about 1.65 kW per square foot, at least about 1.8 kW per square foot, at least about 2.0 kW per square foot, or at least about 3.0 kW per square foot. In some embodiments, the self-sustaining computing system provides computing power of at most about 1.5 kW per square foot, at most about 1.6 kW per square foot, at most about 1.65 kW per square foot, at most about 1.8 kW per square foot, at most about 2.0 kW per square foot, or at most about 3.0 kW per square foot. It will be appreciated that the height of the self-sustaining system can be adjusted to allow more or less computing power to be provided within a given footprint.

[0217] It should be understood that the dimensions, components, arrangements and configurations of the disclosed example embodiments may be modified, increased and / or reduced to produce a variety of potential embodiments in various form factors.

[0218] In some embodiments, a self-sustaining computing system may include a robotic system, such as a gantry robot 526, configured to remove, replace, and / or assemble blade servers, power supplies, or other components, such as chassis 400. A self-sustaining system may include an "internal robotic" or "external robotic" system. In embodiments with a smaller footprint, a smaller chassis 718 for replacing components may be used. In some embodiments, such as Fig.13 As shown, the casing 718 of the replacement parts can be attached to the outside of the tank 710. In some embodiments, the tank 710, the rack, the computer components, the power supply, the replacement casing 718 and the gantry robot 526 can be arranged so that the gantry robot 526 can remove, replace and / or assemble the parts while basically traveling in only one direction. If the various components are arranged substantially linearly, the gantry robot 526 may be able to travel along a single axis to remove, replace and / or assemble the desired parts without traveling in a second direction. It should be understood that in addition to traveling in a single linear direction, the gantry robot 526 may also be able to lift and lower the parts.

[0219] Taking advantage of compact form factors, such as Fig.13 The illustrated embodiment allows for easy transportation of a self-sustaining 2PLIC system. The incorporation of a closed loop chiller 719 allows for the use of a two-phase liquid immersion cooling system in remote conditions where a practical source of chilled water may not be available. Furthermore, the elimination of the need for external chilled water results in a self-sustaining computing system that, in some embodiments, requires only two external connections, a power supply and a data connection.

[0220] In some embodiments, the computing system may be contained within a housing, such as Fig.14In some embodiments, Fig.13 Components schematically identified in and / or disclosed herein may be contained within the housing. In some embodiments, the volume of the housing may be adjusted based on expected cooling requirements, the configuration of the closed-loop chiller, and / or the environment in which the self-sustaining computing system is expected to be deployed.

[0221] The disclosed self-sustaining, self-repairing, compact form factor embodiments can be used as a stand-alone solution to provide significant computing power to nearly any location or environment. In some applications, multiple compact computing systems can be located in close proximity to each other and / or linked together to create a cluster. In some embodiments, the outer shell is arranged to allow maintenance and / or repair operations to be performed with access to only one or both sides of the outer shell. This arrangement allows individual self-sustaining computing systems to be located with a reduced or minimal distance between the individual self-sustaining systems.

[0222] In one example embodiment, a cluster of four exemplary self-sustaining computing systems may be strategically positioned to allow for approximately 240 kW of self-sustaining computer power in a footprint of approximately 140 square feet. In some embodiments, the units may be in electrical and / or data communication with each other, allowing operation of a multi-unit cluster with only a single external power connection and a single data connection. In some embodiments, a data center may be established using multiple compact computing systems or multiple clusters of such computing systems.

[0223] Some disclosed embodiments and / or computing systems disclosed herein can be used in modern data centers and / or climate-controlled environments, however, some embodiments of the disclosed self-sustaining computing systems can be deployed in remote locations and / or harsh environmental conditions. In some embodiments, the outer shell can be weather-resistant, waterproof, and / or otherwise arranged to withstand exposure to harsh environments for extended periods of time. Some disclosed embodiments allow large amounts of computing resources to be quickly deployed to remote or challenging locations. Some self-sustaining embodiments can be arranged to operate in essentially any location with access to power and data connections. In some embodiments, an uninterruptible power supply and / or a generator can be operably connected to the computing system to provide more reliable or consistent access to power.

[0224] Some disclosed self-sustaining embodiments are designed to be stackable. Some stackable embodiments can be designed to have a reduced height. Certain embodiments can be approximately 5'5" tall, 5'6" deep, and 9' wide. This can produce approximately 60kW of computer power in a 42 square foot footprint. Such units can be stacked vertically, providing 120kW of computer power in the same 42 square foot footprint.

[0225] Embodiments of the disclosed computing systems can be stacked, and multiple stacks can be positioned adjacent to each other. This reduces the need for islands of space between individual computing systems, allowing for overall higher power density within a data center.

[0226] Some embodiments may be designed to be fully operable and maintainable by accessing only one side of the self-sustaining computing system. These embodiments may be advantageous because they facilitate positioning self-sustaining systems in close proximity to each other. Additionally, in some self-sustaining embodiments, the entire submerged tank may be removed and / or replaced while accessing only one side of the device. In certain embodiments, the tanks may be individually modular and / or skid mounted.

[0227] In some embodiments, the self-sustaining computing systems can be arranged vertically to utilize an even smaller footprint. A vertically designed embodiment of the disclosed system can provide approximately 60kW of computing power in a footprint of approximately 22.9 square feet. As in some other disclosed embodiments, some vertically oriented self-sustaining computing systems can be positioned in close proximity to each other. Also as discussed in some other embodiments, some vertically oriented self-sustaining computing systems can be operated and maintained by accessing only one side of the device. In some embodiments, the entire tank can be removed and replaced from the outer shell. This arrangement allows for rapid replacement of multiple blade servers and / or other computing components.

[0228] Mobile Embodiment

[0229] A self-sustaining computing system that does not require an external cooling water source allows novel computing applications. In some embodiments, a generator may be used to power the system, thereby removing the need to connect the system to an external and / or fixed power source. In some embodiments, the system may rely on wireless data communications.

[0230] In certain self-sustaining embodiments that do not rely on fixed power sources or wired data communications, a fully mobile computing system can be implemented. The disclosed embodiments include a vehicle-mounted, self-sustaining computing system that can be used to provide significant computing power in almost any environment. In some embodiments, a truck-mounted wireless computing system can be driven within wireless communication range of an existing or ad hoc network and provide large amounts of computing power with essentially no setup or assembly time.

[0231] Examples of natural water

[0232] In some embodiments, the computing system can be arranged to be used on ships, vessels, oil rigs, floating platforms, or other containers or structures positioned close to a body of water. In such embodiments, water from the body of water can be used to cool a condenser, which is used to convert dielectric vapor back into a dielectric liquid, as discussed herein. In an example embodiment, a modular computing system may include a water inlet, a water outlet, and a pump or impeller. The pump and / or impeller may cause water to flow out of the body of water, through the condenser, and then flow into the body of water. Some embodiments may include filters and / or process components that are designed to protect condensers, piping, and other computing system components from potential sources of contamination in the body of water. In some embodiments, condensers and other components are arranged to withstand long-term contact with brackish water or salt water (such as, for example, seawater).

[0233] Horizontal Casing Exchange

[0234] In some embodiments, a case of replacement parts can be stored outside of the case and within the outer shell of the computing system. Replacement parts (such as, for example, chassis, servers, blades, and / or power components) can be removed from the case and used to replace parts within the case. The case can be on a platform that is configured to extend out of the outer shell of the computing system to allow replacement of parts from the case.

[0235] In one non-limiting example, when a blade server in a storage box fails, a robotic arm can be used to extract the non-functioning component from the storage box and move the non-functioning component to a storage slot of the chassis. The robotic arm can then remove a functioning blade server from the chassis and install it in the position where the non-functioning server was previously installed, thereby replacing the non-functioning server with a new functioning server.

[0236] Over time, the casing will accumulate inoperative components that can be replaced by new operational components to allow the robotic system to continue to operate over the long term. In some embodiments, the casing is located on a platform that can extend to the exterior of the outer shell, thereby allowing an operator to access the casing. In some embodiments, the platform is configured to rotate the casing from a substantially vertical position to a substantially horizontal position to allow components to slide into or out of the casing.

[0237] In some embodiments, a height-adjustable cart may be used to move, load, and / or receive components so that a human operator does not need to lift or support the weight of a component when removing or replacing the component from the receiver. It should be appreciated that a receiver configured to rotate to a substantially horizontal position may also facilitate loading properly functioning components into the receiver and removing improperly functioning components.

[0238] Fig.15A-D shows an example casing 810 positioned on a platform 820 that can extend out of the container. Fig.15A In the embodiment, the chassis 810 can be connected to a platform including a rotating member 821, a support member 822, and a track 823. In some embodiments, the support member 822 can be connected to the track 823, which allows the support member 822 to move while supporting the weight of the chassis 810 and any servers or other components stored in the chassis. Fig.15A In the exemplary embodiment of FIG. 8 , platform 820 is in an extended position.

[0239] like Fig. 15B As shown, during normal operation, support member 822 can be retracted relative to the outer shell of the computing system. During normal operation, case 810 can be stored above rail 823. In some embodiments, regardless of the position of support member 822 on rail 823, the weight of case 823 is supported by support member 822 and rail 823.

[0240] In some embodiments, computer components (such as, for example, servers) used with the disclosed embodiments can be denser and / or heavier than conventional computer components. In some embodiments, due to the increased cooling capabilities of the disclosed embodiments, a blade server can weigh at least about 50 pounds, or at least about 60 pounds, or at least about 70 pounds, or at least about 80 pounds, or at least about 90 pounds, or at least about 100 pounds. In some embodiments, a blade server can weigh up to about 50 pounds, or up to about 60 pounds, or up to about 70 pounds, or up to about 80 pounds, or up to about 90 pounds, or up to about 100 pounds. Fig. 15B As shown, the chassis 810 can accommodate multiple chassis 400 or blade servers, wherein a single blade server can weigh up to about 73 pounds. When the chassis is loaded with three such servers, the combined weight of the chassis 810 and the servers can be about 395 pounds.

[0241] In some embodiments, the server used is a blade server mounted on a chassis. The server and / or chassis may include a backplane system to facilitate installation and removal of the server in the computing system. In some embodiments, the server may be an immersion server that does not include fans or other air cooling devices. In some embodiments, a single server board may include 16 GPUs and is configured to draw approximately 6KW of power. In some embodiments, the server is a 1.5U server. Some disclosed servers may be 1 Otto Immersion Unit (Otto Immersion Unit, OIU) servers. Such servers are 1.5U high and are configured for liquid immersion cooling. In some embodiments, a single tank within the computing system may be configured to run ten 1OIU servers and approximately 60KW of power when all ten servers are running at substantially full power. In some embodiments, the computing system may include one or two such tanks. In some embodiments, the computing system may include multiple tanks, such as, for example, ten such tanks.

[0242] In some embodiments, when the case is extracted from the computing system, as Fig.15A As shown, the support member moves along the track from the storage position and is cantilevered to the exterior of the outer shell of the computing system.

[0243] Likewise Figures 15C-15D As shown, the case can be pulled out or otherwise slid along the track and cantilevered to the outside of the computing system. Figures 15C-15D Shown, the casing removal tool can be used to remove the entire casing and the components contained in the casing. In such an embodiment, the casing removal tool can be used to lift the casing from the support member and the slide rails so that the casing can be transported.

[0244] In some embodiments, once the chassis has been moved outside of the computing system, the platform can rotate the chassis to a substantially horizontal position. The servers housed within the chassis can then slide out of the chassis.

[0245] Figures 15A-15D A series of exemplary steps for removing a server from a chassis according to an exemplary embodiment are shown. In an exemplary embodiment, the chassis can be attached to a linear guide rail system behind an access door. Figures 15C-15D As shown, the case can be pulled out and cantilevered to the outside of the computing system. The case can be pulled out manually or moved out of the computing system using a motorized or automated system. Fig.15DAs shown, the case can be rotated approximately 90 degrees to orient the servers and / or other components housed in the case in a substantially horizontal position. Once in the substantially horizontal position, the servers and / or other components can be slid out of the case and onto a cart or other tool configured to receive the servers and / or other components. Fig. 15C As shown, a scissor lift cart can be adjusted to a convenient height to receive a server or other component. An adjustable height cart with a rolling surface can be used to allow the server to be transferred from the chassis to the cart without requiring a human operator to support the weight of the server. Fig.15D As shown, once the server slides onto a cart with a sliding or rolling surface, the server or other components can be transported to other locations for replacement or repair. It is understood that new components can be loaded into the cassette using substantially the same steps in reverse order.

[0246] In some alternative embodiments, the case can be supported on a rotating and extendable arm without a track. In such an embodiment, during normal operation, the case can be stored in a substantially vertical position within the outer shell of the computing system. Once it is determined that the components in the case should be replaced, the extendable arm can be used to extend the case to the outside of the outer shell. Once the case extends beyond the outer shell, the case can be rotated from a substantially vertical position to a substantially horizontal position to allow the components stored in the case to be removed horizontally from the case.

[0247] Bellows

[0248] In certain embodiments, bellows and / or steam collection system can be used.Before some disclosed embodiments are initially enabled, dielectric fluid, computer components such as servers and other system components can be in thermal equilibrium.Once computing system is enabled, computer components such as servers may begin to generate heat, and the heat can be distributed in the dielectric fluid.This process causes some dielectric fluids to be transformed into vapor state from liquid state.Along with the temperature rise of fluid, a larger proportion of dielectric fluid can be transformed into vapor state.In closed system, the volume of the increase of dielectric steam may cause the pressure of the increase in the system.In certain embodiments, the tank comprising dielectric fluid can be communicated with recovery system fluid and / or steam.

[0249] Fig.16A vapor recovery system 900 is shown according to an example embodiment. The recovery system 900 is connected to a tank 710 containing dielectric vapor. The dielectric vapor can flow from the tank 710 through a pipe to one or more bellows 905. In some embodiments, the vapor recovery system 900 includes an expanding and contracting bellows 905 configured to receive the dielectric vapor, thereby reducing or eliminating any pressure buildup in the tank 710. When the system cools or a portion of the dielectric vapor condenses into a dielectric liquid, the bellows can contract or retract to substantially maintain pressure balance within the tank 710.

[0250] In some embodiments, the vapor recovery system 900 includes a valve 912 configured to allow ambient air to enter the vapor recovery system. In such embodiments, the dielectric vapor can be mixed with the ambient air. Mixing the dielectric vapor with the ambient air can reduce the temperature of the dielectric vapor. In some embodiments, the mixed air / steam can be directed through the carbon bed 911. The carbon media within the carbon bed 911 can be configured to attract the dielectric vapor while allowing ambient air to pass through the carbon media and exhaust from the system 900, such as through an outlet valve 913. In this way, the heated dielectric vapor can be cooled and captured by the carbon media.

[0251] After running for a sufficiently long time, embodiments of the computing system will reach a stable thermal state based on the power capacity of the computing components used. If more or less computing power is used, more or less dielectric fluid may be converted to dielectric vapor. This may cause bellows 905 to expand and / or contract in response to the heat dissipated into the dielectric fluid.

[0252] In some embodiments, the bellows 905 may include one or more storage bags. Each storage bag may include a metal foil and a polymer laminate structure. The bellows storage bags may be connected to each other and the vapor recovery system piping in series or in parallel. In some embodiments, the total volume of the expanded bellows storage bags may be at least about 15% of the liquid fluid volume of the tank. In some embodiments, the total volume of the expanded bellows storage bags may be at least about 20% of the liquid fluid volume of the tank, or at least about 23%, or at least about 25%, or more. In some embodiments, the total volume of the expanded bellows storage bags may be at most about 40% of the liquid fluid volume of the tank, or at most about 30%, or at least about 25%, or less.

[0253] In some embodiments, when the computing system has substantially reached thermal stability, the vapor recovery system 900 can be closed to cool the ambient air, and the valve that allows air to be discharged outside the system can be closed. In some embodiments, the carbon bed can be configured to open only to the tank and bellows using a valve. In some embodiments, a desorption heater configured to transfer heat to the carbon medium can be enabled to raise the temperature of the carbon medium. Along with the rising of the temperature of the carbon medium, any dielectric fluid previously captured by the carbon medium can be driven out of the carbon and returned to the tank, where, as previously described, it can be condensed back into the dielectric fluid.

[0254] In some embodiments, when the computing system is powered at a lower level than the previous stable state, the portion of the dielectric fluid in the vapor state may decrease, and in some embodiments, the bellows may retract to accommodate the reduction in dielectric vapor. In some embodiments, a valve that allows ambient air to enter the bellows may be opened to allow air to enter the bellows and further reduce any pressure differential. In some embodiments, nitrogen may be used instead of ambient air to reduce the pressure differential and also avoid introducing any potential contaminants from ambient air.

[0255] In some embodiments, the bellows and / or vapor recovery system can be completely or substantially passive. In some embodiments, the bellows and / or vapor recovery system can be powered and / or automated based on sensor data from temperature sensors, pressure sensors, and / or power sensors positioned throughout the computing system.

[0256] In some embodiments, a computing system with a vapor recovery system is emission-free, even if the system is not a closed system. In some embodiments, ambient air or nitrogen can be introduced into the system and exhausted from the system without releasing any or substantially any dielectric fluid into the surrounding atmosphere.

[0257] Example Embodiments

[0258] The disclosed embodiments allow for increased density of computer components and / or computing power. In some embodiments of computer components 170 comprising two-phase liquid immersion cooling within a controlled pressure vessel 110, the components can be separated from each other by less than about 1", or less than about 0.7 inches, or less than about 0.5 inches. In some embodiments, individual components can be separated by greater than about 0.3 inches, or greater than about 0.5 inches, or greater than about 0.7 inches, or greater than about 1 inch, or greater than about 1.5 inches.

[0259] Some disclosed embodiments allow for improved power utilization effectiveness (PUE) compared to conventional data centers. Use of the disclosed embodiments allows for reduced energy usage for cooling computer components 170, thereby reducing overall energy usage for the data center and bringing the PUE closer to 1.0. Some embodiments relate to a data center including computer components with two-phase liquid immersion cooling within a controlled pressure vessel 110, wherein the data center has a PUE of less than about 1.15, or less than about 1.10, or less than about 1.08, or less than about 1.05. Some embodiments relate to a data center including computer components with two-phase liquid immersion cooling within a controlled pressure vessel 110, wherein the data center has a PUE greater than about 1.05, or greater than about 1.06, or greater than about 1.08, or greater than about 1.10.

[0260] In certain embodiments, heat-conducting condensable dielectric fluid is provided for two-phase liquid immersion cooling system.Computer components are operated at a pressure lower than the ambient atmospheric pressure, which reduces the temperature of dielectric fluid evaporation, thereby maintaining the liquid phase of dielectric fluid at a lower temperature compared with standard atmospheric pressure.Computer components can generate heat when running.The heat generated is transferred to the dielectric liquid in contact with the computer components, causing the dielectric liquid to evaporate into gas.Gaseous dielectric fluid can use a condenser to condense.Ambient temperature or freezing process water passes through the condenser.When the gaseous dielectric fluid is cooled by the condenser, it condenses back to the liquid phase and falls back into the bath of the liquid dielectric fluid.

[0261] Some disclosed embodiments relate to high-density data centers. Conventional data centers include about 1 megawatt (MW) of computing power distributed over about 10,000 square feet. High-end data centers may include about 1 MW of computing power distributed over about 6,000 square feet. The disclosed embodiments relate to a data center including two-phase liquid immersion cooled computer components 170 within a pressure controlled vessel 110, wherein the data center utilizes about 1 MW of computing power distributed over about 3,000 square feet, or about 1,500 square feet, or about 1,000 square feet, or about 800 square feet, or about 600 square feet. In some embodiments, multiple pressure controlled vessels containing the disclosed computing system can be arranged in a row and powered by a central power supply. In some embodiments, multiple embodiments of the disclosed computing system can be connected in series with each other.

[0262] The disclosed embodiments include liquid immersion cooled computer components 170 within a controlled pressure vessel 110, whereby the components are isolated from atmospheric contaminants by the controlled pressure vessel and immersion in a dielectric liquid 140. Some disclosed embodiments relate to data centers that operate with minimal air filtration and / or cleaning requirements. In some embodiments, the data center operates without HEPA filters or equivalent, or without MERV 11 filters or equivalent, or without MERV 8 filters or equivalent.

[0263] The disclosed embodiments include liquid immersion cooled computer components 170 within the controlled pressure vessel 110, so these components are not cooled by gaseous air. The disclosed embodiments include data centers that operate without cooling fans and / or any other similar means for circulating air.

[0264] The disclosed embodiments relate to an environmentally friendly data center. In some embodiments, the data center includes a computer component 170 of liquid immersion cooling in a pressure controlled container 110, and consumes little or no water for the cooling process. Some embodiments utilize a closed loop dry cooling tower to reduce the temperature of the water circulating through the disclosed condensing structure 130, so as to cool the condensing structure 130 and condense the dielectric fluid vapor into a dielectric fluid liquid. Such embodiments operate without significant water input or output because the closed loop dry cooling tower does not rely on water flow or evaporative cooling for cooling operations. Some data center embodiments utilize and / or discharge less than about 10,000 gallons of water per day, or less than about 1,000 gallons of water per day, or less than about 100 gallons of water per day, or less than about 10 gallons of water per day, or 0 gallons of water per day. Some data center embodiments utilize and / or discharge more than about 100 gallons of water per day, or more than about 1,000 gallons of water per day, or more than about 10,000 gallons of water per day.

[0265] The disclosed embodiments relate to a computing system comprising: a pressure-controlled vessel operably connected to a pressure controller and / or a vacuum source, wherein the pressure-controlled vessel has an interior and an exterior and is configured to contain a certain atmosphere in the interior; a certain volume of thermally conductive, condensable dielectric fluid; a rack for mounting computer components, wherein the rack is arranged so that when the computer components are mounted on the rack, the computer components are at least partially immersed in the volume of thermally conductive dielectric fluid; and a condensing structure, wherein the volume of thermally conductive dielectric fluid, the rack, the computer components and the condensing structure are contained within the pressure-controlled vessel. Some embodiments relate to a cooling system comprising: a pressure controlled vessel including an interior, wherein the vessel is configured to be operably connected to a pressure controller to reduce the internal pressure to below atmospheric pressure, wherein the pressure controlled vessel is configured to include a volume of a thermally conductive, condensable dielectric fluid in a liquid phase and a vapor phase; one or more computer components arranged so that the one or more computer components can be at least partially immersed in the liquid phase of the volume of the thermally conductive, condensable dielectric fluid; and a condenser for condensing the vapor phase dielectric fluid into a liquid phase dielectric fluid.

[0266] In some embodiments, a pressure controlled container is mounted within a super structure, the blade servers are configured to be swappable without disrupting the computing system, the pressure controlled container is operably connected to a power source, a water source, and a network connection, the pressure controlled container includes an opening on the top and a lid configured to sealably close the opening, the lid configured to direct rising steam from the middle of the pressure controlled container to the side of the pressure controlled container, the pressure controlled container has an internal volume between about 100 cubic feet and about 300 cubic feet, and / or wherein the pressure controlled container contains a ratio of liquid dielectric fluid to gaseous dielectric fluid between about 1:3 and about 1:8. Some embodiments also include a ballast block, blade servers and blade server chassis, a robotic arm, and an airlock, wherein the airlock is configured to allow access to the interior of the pressure controlled container without significantly disrupting the atmosphere within the pressure controlled container, and / or a purge system, wherein the purge system is configured to remove contaminants from the volume of the thermally conductive dielectric fluid. In some embodiments, the purification system is configured to remove a portion of the atmosphere from the pressure controlled container, condense any dielectric fluid from the atmosphere, and discard any remaining vapor. In some embodiments, the purification system is configured to condense at least a portion of the gaseous dielectric fluid and discard the gaseous contaminants.

[0267] Some embodiments relate to a method for cooling computer components, the method comprising: providing a housing, wherein the housing contains a heat-conducting, condensable dielectric fluid and a heat-generating computer component, and wherein the housing is configured to withstand at least a slight vacuum; operating the computer component, wherein operating the computer component generates heat, and wherein the computer component is in contact with the dielectric fluid; and generating a vacuum in the housing, wherein the pressure in the housing is at least less than about 1 atmosphere. Some embodiments also include maintaining a vacuum in the housing, wherein when the computer component is running, the pressure in the housing is less than about 1 atmosphere, using the heat generated by the computer component to evaporate the dielectric fluid from a liquid state to a gaseous state, and using a condenser to condense the dielectric fluid from a gaseous state to a liquid state, thereby removing fluids that are not easily condensed from the dielectric fluid. In some embodiments, and / or replacing a portion of the computer component while the system is running. In some embodiments, removing non-condensable fluids includes: isolating a portion of the gaseous atmosphere from the atmosphere in the housing, condensing any dielectric fluid from the gaseous atmosphere; returning the condensed dielectric fluid to the housing, and discarding any remaining portion of the gaseous atmosphere, and / or the housing is configured to generate convection.

[0268] Some embodiments relate to a method of cooling a computer component, comprising operating the computer component at a pressure below ambient, wherein the computer component is in contact with a thermally conductive dielectric fluid. Some embodiments also include evaporating the dielectric fluid and condensing the dielectric fluid at a pressure below ambient.

[0269] Some embodiments relate to a method for cooling a computer component, the method comprising: providing a thermally conductive, condensable dielectric fluid in a liquid phase and a vapor phase; and operating the computer component at a pressure below ambient atmospheric pressure in the presence of the thermally conductive, condensable dielectric fluid, wherein the computer component is at least partially in contact with the thermally conductive, condensable dielectric fluid in the liquid phase. Some embodiments also include using at least a portion of any heat generated by the operating computer component to evaporate the dielectric fluid from a liquid phase to a vapor phase; condensing at least a portion of the dielectric fluid from a vapor phase to a liquid phase; removing at least a portion of the non-condensable fluid from the dielectric fluid; and / or replacing at least one or more computer components while the computer component is operating.

[0270] Some embodiments are directed to a method of cooling a computer component, the method comprising: operating the computer component at a pressure of at least 1 psi below ambient pressure, wherein the computer component is at least partially in contact with a thermally conductive dielectric fluid, and wherein the boiling point of the dielectric fluid is less than about 80° C. Some embodiments further comprise condensing the dielectric fluid at a temperature of no greater than about 80° C. at the computer component.

[0271] It will be appreciated that the various disclosed embodiments may be incorporated into some or all of the components otherwise described herein. Specific components and their properties may be adjusted based on the properties of each specific embodiment. Modifications may include the use of higher or lower density power, cooling and network connection systems, pressure management systems, steam management systems, and the selection of special equipment and components.

[0272] From the foregoing description, those skilled in the art can easily determine the basic essential features of the present disclosure, and without departing from its spirit and scope: various changes and modifications can be made to adapt the present disclosure to various uses and conditions. The embodiments described above are meant to be illustrative only and should not be regarded as limiting the scope of the present disclosure.

[0273] Heating and cooling of storage tanks in response to shock events

[0274] In an exemplary embodiment, the immersion cooling system or container may include a tank, a computing device, a robot, an absorption unit, a bellows, and a management system. The tank may be a pressure-controlled tank maintained at atmospheric pressure (or within its range). The tank may include a bathtub area and a liquid collection area, and the computing device may be immersed in the dielectric fluid in the bathtub area of ​​the tank. The computing device may be connected to a network and perform various processing tasks while being immersed in the dielectric fluid. The tank may include a lid for entering the bathtub area, the computing device, and the liquid collection area. The tank may be fluidically coupled to the bellows and the absorption unit, and a plurality of valves may selectively connect the tank to the bellows and the absorption unit or disconnect the tank from the bellows and the absorption unit so that dielectric vapor may be transferred to the bellows and / or the absorption unit, or vice versa. The robot may be a gantry robot that may lift the computing device from the tank of the container when the lid of the tank is opened. The robot may place the lifted computing device in a casing provided for storing the computing device. The robot may also lift the computing device from the chassis and place it in the location of the computing device lifted from the tote.

[0275] In an example embodiment, the tank may include a heating element, such as a plurality of heating rods, some of which are at least partially immersed in the dielectric fluid. The tank may include a plurality of sensors, such as a temperature sensor, a pressure sensor, or a sensor providing operational data (such as current, voltage, workload, etc.) associated with the computing device. The temperature sensor may be located in an area inside the bath or above the bath. Using the data received from the sensor, the management system of the container may operate the heating element to regulate or control the temperature or temperature fluctuation (and / or the pressure or pressure fluctuation of the dielectric steam) of the dielectric fluid in the tank. Fig.18An example of a heating element 1000 for an immersion cooling system according to an exemplary embodiment is shown. The heating element 1000 may include a plurality of heating rods 1010. Each heating rod may include a plurality of wires 1011 that may be connected to a power source of the tank. The controller of the tank may adjust the heating element 1000 to heat a bath area of ​​the tank, such as during various operations of the tank. In this exemplary embodiment, the heating element 1000 may be mounted in a bath of the tank and completely immersed in the dielectric fluid.

[0276] In one example embodiment, the heating element is separate from the computing device and the heating element does not process data. The heating element may be dedicated to generating heat only and has no other functions. The heating element may be easily controlled, particularly during operation of the tank (e.g., startup operation), replacement of components, or other times when control is desired. The heat generated by the heating element may be adjusted while evaluating data indicating the size of the bellows and other aspects of the system (e.g., pressure or temperature).

[0277] Specifically, the rapid change of the power consumption or workload of computing equipment (for example, caused by end-user activity or lack of activity) may cause the rapid change of the amount of heat generated by the computing equipment in the container. This and then may cause the rapid temperature change in the tank or bath, thereby may cause the sudden change of the pressure in the tank (because in a closed insulation system, pressure and temperature are directly related, i.e. PV=nRT). These pressure fluctuations may damage the container and guide polluted gas (such as air) or particulate (such as dust) into the tank. These pressure fluctuations may also cause dielectric fluid to leak from the tank. In order to offset the influence of these pressure fluctuations, when the pressure drops, bellows or absorption unit can be used to remove excessive steam from the tank or steam is introduced into the tank. However, by using a heating element, the capacity of the bellows and absorption unit can be reduced, thereby a more space-saving container can be designed. If a heating element is not used, the bellows may break when the pressure is excessively increased.

[0278] The heating element can allow the temperature change within the tank or bath to be regulated, thereby facilitating the controlled transition between various operating load conditions that the computing device may experience during its operation. For example, in the case of a rapid reduction in the operating workload of the computing device, the heat generated by the computing device can drop rapidly. This may cause the internal pressure of the tank to drop suddenly. The heating element can add heat to the tank to achieve a controlled reduction in the temperature of the dielectric fluid, such as during a shutdown process. In other words, in the case of a sudden change in the workload of the computing device, i.e., a shock event, the heating element can balance the pressure and temperature of the tank. Therefore, the container requires a much smaller bellows and absorption unit to maintain the atmospheric pressure of the tank.

[0279] In an exemplary embodiment, the management system of the container can be in response to an impact event (such as an increase or reduction of the internal pressure or temperature of a tank), to determine how much heat to add to the tank. In an exemplary embodiment, the rate of the decline (or rise) of temperature or pressure can determine how much heat to add to the tank. For example, when the dielectric fluid temperature level of the tank is reduced to exceed a certain degree, in a certain number of minutes, the management system can enable a heating element to add a certain amount of heat (such as to keep the temperature and pressure of the system) to the tank. The heat of this addition can cause the temperature to stop declining or the rate of temperature decline to reduce. When the tank is in a stable state, for example, when the rate of decline of pressure or temperature is lower than a threshold value, the management system can stop the heating element to add heat to the system. In another exemplary embodiment, when the increase or reduction of the workload of a computing device begins, the actual temperature of the dielectric fluid in the tank can determine how much heat to add to the tank.

[0280] In an exemplary embodiment, the management system can activate the heating element when a shock event is detected, such as before, during, or after a startup operation, a load increase operation, a load reduction operation, or a shutdown operation. The management system can detect the operating mode of the container (e.g., startup or shutdown) by receiving sensor data (e.g., from a temperature or pressure sensor in the tank) or data from a computing device (e.g., current, voltage, temperature, workload, data transmission, etc.). The heating element can adjust or regulate changes in temperature or pressure within the tank to minimize deviations in pressure from atmospheric pressure. Otherwise, without operating the heating element according to the techniques disclosed herein, the container will have to absorb or store excess gas generated by the rapid heating of the computing device, or the container will have to desorb or supply gas to offset the pressure drop caused by the rapid drop in heat generation of the computing device.

[0281] During the startup operation, when the computing device starts operation, the temperature of the tank (such as the temperature of the dielectric fluid in the bath) is lower than the threshold value. The startup operation can occur, for example, when the container has just been opened and the tank is cold. Because the computing device can heat quickly, if the dielectric fluid is cold, the computing device can produce a large amount of steam. Therefore, before, during or after the startup operation, the management system can start the heating element to heat the dielectric fluid, thereby increasing the temperature of the dielectric fluid in a controlled manner, and minimize the steam generated by the computing device. For example, before the computing device is turned on, the heating element can slowly increase the temperature of the dielectric fluid to the threshold temperature. Otherwise, the container will have to accommodate excessive steam, so that the tank is maintained at atmospheric pressure, which may require large capacity at bellows and absorption unit.

[0282] During the load-up operation, the temperature of the tank, such as the temperature of the dielectric fluid in the bath, can increase faster than the threshold rate (for example, when the temperature of the tank is below the threshold). For example, when computing equipment is running and the workload of computing equipment increases significantly (for example, due to the increase of consumer demand), the load-up operation can occur. The sudden increase of the workload of computing equipment may increase the amount of heat generated by the computing equipment, thereby increasing the amount of the steam generated by the computing equipment. Therefore, before, during or after the load-up operation, the management system can enable the heating element to heat the dielectric fluid, thereby increasing the temperature of the dielectric fluid in a controlled manner, and minimizing the steam generated by the computing equipment. Otherwise, the container will have to accommodate excess steam, to keep the tank under atmospheric pressure, which may require large capacity for storage or absorption at bellows and absorption unit.

[0283] During the load reduction operation, the temperature of the tank, such as the temperature of the dielectric fluid in the bath, can be reduced faster than the threshold rate (for example, when the temperature of the tank is higher than the threshold value). For example, when the computing device is running and the workload of the computing device is significantly reduced, for example, due to a decline in consumer demand, a load reduction operation may occur. The sudden reduction of the workload of the computing device can reduce the amount of heat generated by the computing device, thereby suddenly reducing the pressure in the tank. Therefore, before, during or after the load reduction operation, the management system can enable a heating element to heat the dielectric fluid, thereby reducing the temperature of the dielectric fluid in a controlled manner, and minimizing the pressure drop in the tank. Otherwise, the container will have to produce a large amount of steam, to keep the tank under atmospheric pressure, which may require large storage or desorption capacity at bellows and absorption unit.

[0284] During the closing operation (or controlled closing process), when the temperature of the tank (such as the temperature of the dielectric fluid in the bath) is higher than a threshold value, the container is ordered to close. Because the computing device can suddenly stop generating heat, the pressure in the tank can drop quickly. Therefore, before, during or after the closing operation, the management system can enable the heating element to heat the dielectric fluid, thereby reducing the temperature of the dielectric fluid in a controlled manner, and minimize the pressure drop. For example, the heating element can slowly heat the dielectric fluid so that when the computing device is closed, the temperature of the dielectric fluid slowly decreases. Otherwise, the container will have to produce a large amount of steam, so that the tank is maintained under atmospheric pressure, which may require large storage or desorption capacity at bellows and absorption unit.

[0285] In one example embodiment, as the vessel responds to a shock event, the management system (or another system) can add steam or fluid to the tank, or remove steam or fluid from the tank, to maintain the pressure of the tank at a pressure close to atmospheric pressure. For example, as the temperature of the tank increases, steam or fluid can be removed from the tank, and as the temperature of the tank decreases, steam or fluid can be added to the tank.

[0286] The container can use various mechanisms to add steam or fluid to the tank, or remove steam or fluid from the tank. In an exemplary embodiment, the container can use a bellows as a mechanism for adding steam to the tank or removing steam from the tank. In another exemplary embodiment, the container can use an absorption / desorption unit (hereinafter referred to as an "absorption unit") to add steam to the tank or remove steam from the tank. In yet another exemplary embodiment, the container can use a pressurized vessel to add steam to the tank or remove steam from the tank. In yet another exemplary embodiment, the container can use a combination of the above mechanisms to add steam or fluid to the tank or remove steam or fluid from the tank. In yet another exemplary embodiment, the container can use a combination of a heating element and one or more of the above mechanisms to maintain the pressure of the tank.

[0287] For example, during the startup operation, the management system can use a combination of a heating element and a bellows to maintain the pressure of the tank. In one example, before the computing device is turned on, the management system can start the heating element to heat the dielectric fluid. At a certain point (e.g., before, after, or during heating), the management system can open the valve connecting the bellows to the tank, thereby promoting the transfer of dielectric steam to the bellows. This transfer of dielectric steam to the bellows can prevent the uncontrolled increase of the pressure of the tank, thereby allowing the temperature of the dielectric fluid to increase, while the pressure of the tank can be maintained (e.g., within a tolerance range).

[0288] Similarly, during startup operations, the management system can use a combination of a heating element and an absorption unit to maintain pressure in the tank. At some point (e.g., before, after, or during heating), the management system can open a valve connecting the absorption unit to the tank, thereby facilitating transfer of dielectric vapor to the absorption unit, which can absorb or maintain dielectric vapor in the absorption unit (e.g., a carbon bed). Similarly, during startup operations, the management system can use a combination of a heating element and a pressurized vessel to maintain pressure in the tank. At some point (e.g., before, after, or during heating), the management system can open a valve connecting a pump and a pressurized vessel to the tank, thereby facilitating transfer of dielectric vapor to the pressurized vessel using a pump. The pressurized vessel can store dielectric vapor.

[0289] As another example, during the shutdown operation, the management system can use a combination of a heating element and a bellows to maintain the pressure of the tank. In one example, after shutting down the computing device, the management system can start the heating element to heat the dielectric fluid. At a certain point (e.g., before, after, or during heating), the management system can open the valve connecting the bellows to the tank, thereby facilitating the transfer of dielectric vapor to the tank. This transfer of dielectric vapor to the tank can prevent the pressure of the tank from being uncontrolledly reduced, thereby allowing the temperature of the dielectric fluid to be reduced, while the pressure of the tank can be maintained (e.g., within a tolerance range).

[0290] Similarly, during shutdown operations, the management system can use a combination of a heating element and an absorption unit to maintain the pressure of the tank. At some point (e.g., before, after, or during heating), the management system can open a valve connecting the absorption unit to the tank, thereby facilitating the transfer of dielectric vapor to the tank. In the case of a carbon bed as the absorption unit, the management system can activate the carbon bed to release the trapped or absorbed dielectric molecules. The management system can activate the carbon bed by, for example, sending a signal to a switch to turn on a heating device within the carbon bed. In one example, as the pressure of the tank drops, the carbon bed can be heated to release the dielectric vapor and minimize the drop in pressure.

[0291] Similarly, during shutdown operations, the management system can use a combination of a heating element and a pressurized vessel to maintain the pressure of the tank. At some point (e.g., before, after, or during heating), the management system can open a valve connecting the pressurized vessel to the tank, thereby facilitating the transfer of dielectric vapor to the tank.

[0292] In an example embodiment, a trade-off may be made between using bellows and absorption cells. Bellows are passive devices, but absorption cells are active devices. Systems based on bellows are more energy efficient than systems based on absorption cells because the bellows do not require active heating. However, bellows take up more space than absorption cells, and absorption cell-based systems offer a greater degree of control and functionality. Some design constraints in this regard may be efficiency, control, and space.

[0293] In one example embodiment, the container may experience an uncontrolled shutdown. For example, the container may experience an uncontrolled shutdown due to a power outage. In this example embodiment, an emergency shutdown process may be implemented to address possible pressure fluctuations in the tank. For example, the container may have a backup or uninterruptible power supply ("UPS") that can supply power to the container and its management system (or another system). If the management system receives a signal from a sensor indicating that the pressure in the tank has dropped below an acceptable threshold (caused by system power outages and cooling), the management system can instruct the bypass valve to open. The bypass valve can connect the tank to the environment outside the tank. The bypass valve can introduce air into the tank and thereby normalize the pressure in the tank (so that the tank or bellows will not collapse). Subsequently, during the startup operation, the container can purge the air introduced into the tank.

[0294] In one example embodiment, a management system (or another system) may use a table, matrix, or map ("map") to determine how to respond to a shock event. In one example embodiment, the map may display a change in temperature as an input and display how much heat is added to the tank in response to the change in temperature as an output. In one example embodiment, the map may include data related to steam temperature, tank pressure, fluid level in the tank or sump area, fluid pressure in a pump or filter, pressure differential, humidity level, and aluminum oxide condition as inputs. In response to these inputs, the map may provide outputs such as condensers, heating elements, pumps, bellows valves, carbon intake valves, carbon exhaust valves, and computing device operating parameters. The map may define various states of operation of the container. The management system may receive various data from sensors that are disposed throughout the container. Using the map, the management system may convert the data into operating parameters for devices on the container (e.g., bellows, absorption cells, valves, heating elements, pumps, condensers, and computing devices).

[0295] In one example embodiment, the container can be operated at a temperature close to the boiling point of the dielectric fluid and a pressure close to atmospheric pressure. However, one of ordinary skill in the art recognizes that the container can be operated at other temperature and pressure ranges based on the optimal operating temperature of the operating computing device. In one example embodiment, the optimal operating temperature of the system is about 137℉±8°. In one example embodiment, the optimal operating pressure of the system is about atmospheric pressure (e.g., 101,325Pa)±5,000Pa. In this example embodiment, during a shock event, the management system will attempt to keep the temperature and pressure of the container within these ranges.

[0296] Although in some example embodiments of the present disclosure, the management system is specified as a system programmed to perform various tasks in a shock event, one of ordinary skill in the art recognizes that other systems disclosed in the present disclosure may be programmed to perform these tasks.

[0297] In one example embodiment, the container can operate in three operating modes. In the first operating mode, the tank can operate at atmospheric pressure. In the second operating mode, the tank can operate at a pressure range that significantly deviates from atmospheric pressure. In the third operating mode, the container can sometimes operate at atmospheric pressure and sometimes operate at a pressure range that significantly deviates from atmospheric pressure. The third operating mode can be a mixture of the first mode and the second mode. In an example embodiment, the management system can determine the operating mode of the container. For example, the management system can operate the container based on the rules defined for the management system, for example, pressurizing the container at 5AM every morning and returning to atmospheric pressure at night; pressurizing the container during the peak workload determined by the sensor data. As another example, the management system can use a machine learning algorithm to predict the operating mode of the container. For example, the machine learning algorithm can use sensor data as well as external data (such as weather conditions, calendar data, usage data, etc.) to predict which operating mode is more effective in this case. Users of the system can provide tagged data to the management system, and the management system can extrapolate the data to create a model for predicting the operating mode.

[0298] In one example embodiment, the management system may perform a specific routine before the lid of the tank can be opened. For example, if the container is provided with instructions to open the lid of the tank, the condensation system may cool the system for a period of time before the management system allows the lid to be opened. The condensation system may minimize vapors in the tank so that when the lid is opened, minimal dielectric vapor is lost to the environment.

[0299] In one example embodiment, the immersion cooling system can be a modular system. For example, each set of components of the system can be mounted on a separate skid, such as a condensing skid, a heating skid, a bellows skid, an absorption unit skid, etc. These skids can be removable and can be used for various applications.

[0300] Dielectric fluid circulation and filtration

[0301] In an exemplary embodiment, the container may include a pump for circulating the dielectric fluid through the tank. For example, the tank may include a liquid collection area and a bathtub area. The bathtub area may keep a computing device immersed in the dielectric fluid. The liquid collection area may be near the bathtub area, or the liquid collection area, and may be in fluid communication with the bathtub area. For example, the liquid collection area may receive an overflow of the dielectric fluid from the bathtub area, for example, the dielectric fluid may flow through the wall of the bathtub area adjacent to the liquid collection area. The pump may extract the dielectric fluid from the liquid collection area and pass the fluid through a filter. After the filter, the dielectric fluid may return to the bathtub area. The container may include various pipelines connecting the liquid collection area, the pump, the filter, and the bathtub area.

[0302] In an example embodiment, the container is provided with a certain amount of dielectric fluid so that the bathtub area is full of dielectric fluid, and there is an overflow of dielectric fluid in the liquid collection area. The bathtub area full of guarantees that the computing device is fully immersed in the dielectric fluid. The pump can extract the dielectric fluid from the liquid collection area, and deliver it to the bathtub area, for example, by a filter. Because the dielectric fluid in the storage tank has more capacity than the bathtub area to accommodate fluid, when the pump is running, the bathtub area is always full of (particularly when the pump is operated). However, according to the temperature of the storage tank, the liquid level of the dielectric fluid in the liquid collection area may change, because the dielectric fluid may evaporate from the bathtub area, and the dielectric fluid from the liquid collection area may replace the fluid evaporated in the storage tank.

[0303] In an example embodiment, the tank can be a rectangular box shape. The dielectric fluid can flow through the top of one of the shorter sides and enter the liquid collection area adjacent to the shorter side. The pump can extract the dielectric fluid, and return it or reintroduce it into the position in the tank, which can cause minimal interference or turbulence to the fluid in the bathtub, because interference or turbulence may cause cavitation in the fluid. Especially, the overflow area is longer to the distance between the reintroduction point, and the turbulence related to the reintroduction of fluid in the tank is less. For example, if the dielectric fluid overflows from the top of the first side of the bathtub, the pump can make the dielectric fluid return to the bottom of a side opposite to the first side. The pump can return the dielectric fluid to the corner of the bottom side, which minimizes interference or turbulence of fluid in the bathtub.

[0304] In one example embodiment, the container may include two pumps. Each pump may independently draw fluid from the sump area and deliver it to the bath area. Providing the container with two separate and independent pumps may increase the useful life of the container. Furthermore, if one of the pumps fails for any reason, the container may continue its operation without interruption until the failed pump is replaced.

[0305] In an exemplary embodiment, the container may include a filter. The filter may include one or more cores. Each core may filter different types of pollutants, particles, substances, diluents or solutes in the dielectric fluid. In an exemplary embodiment, the core may be selected based on the properties of the dielectric fluid and the pollutants that may be introduced into the dielectric fluid. For example, the pollutants may include solder and resin, which are used in the process of manufacturing the electronic board used in the computing device. The dielectric fluid may be used as a cleaning agent for resin, solder, dust, dirt or any other substance in the system. Solder and resin (or other substances) may be washed off from these electronic boards after the electronic boards are immersed in the dielectric fluid. The filter may remove solder and resin (or other substances) from the dielectric fluid. If these substances are not removed from the dielectric fluid, when the dielectric fluid evaporates, these substances will be deposited into layers on the heat generating components (such as processors) of the computing device. Therefore, the layer thermally isolates or thermally insulates the heat generating components from the dielectric fluid, thereby reducing the heat transfer efficiency from these components to the dielectric fluid. Therefore, the components may heat up and be damaged more frequently.

[0306] In one example embodiment, the filter may include two cores, one core including activated carbon (charcoal) and one core including activated aluminum. For example, the ratio of activated carbon to activated aluminum may be 3 to 1. As another example, the filter may include four cores, three cores including activated carbon and one core including activated aluminum.

[0307] In an example embodiment, the filter may include a strip for testing the acidity of the dielectric fluid. The strip may be a pH indicator, litmus paper or other indicator. In some cases, the dielectric fluid may become acidic after interacting with certain components of the tank. If the dielectric fluid becomes acidic, the strip may contact the dielectric fluid and change color. The filter may include a color detection sensor that may detect the color change of the strip, and if the color change of the strip is detected, a signal is sent to a management system (or another system). In an example embodiment, the strip may be arranged in a vessel or chamber including a glass shield. Therefore, the change in strip color is visible outside the vessel. A camera may be arranged near the vessel. The camera may take a photo of the strip (behind the glass shield) and send the photo to the management system. If the management system (or a user of the system) detects a change in strip color (using data provided by a camera or color sensor), the management system may trigger a remedial action, such as notifying a maintenance system or shutting down the system.

[0308] In one example embodiment, the camera can be a pan-tilt-zoom camera. A filter cover can be mounted on top of the liquid collection area. The filter cover can be mounted adjacent to other covers that provide access to the bath area. The filter cover can include a filter and the camera can be mounted on the filter cover. In one embodiment, the camera can be mounted directly below the filter cover. In this way, as the camera rotates, the camera can capture images of the strip, the liquid collection area (the area below the camera), and the area above the bath area.

[0309] Fig.19A -B shows a filter including three cores according to an example embodiment. Fig.19A As shown, the filter can include a cover 1050 that can be mounted on the tank, for example, adjacent to other covers that provide access to computing devices mounted in the tank. Each core of the filter can be connected to the cover 1050. The cover 1050 can include three caps 1060, each cap providing access to a core. Fig.19B A structure 1070 is shown mounted on the cover 1050. The structure 1070 can support various filter cartridges and other components, such as a filter cartridge 1071, a camera 1072, and an electromechanical valve 1073. On the other side of the structure 1070, there can be two more filter cartridges ( Fig.19B not shown).

[0310] In this example filter, there is a camera and two color sensors attached to the cover. The camera and color sensors can obtain data related to the acidity of the dielectric fluid (based on the color of the strips) and transmit the data to the management system.

[0311] In one example embodiment, the filter may be mounted on a chassis that can be removed by the robot. The chassis may include a connection interface for removably connecting the chassis (and the filter disposed therein) to various conduits disposed in the tank. Thus, when the management system determines that the filter needs to be replaced, the robot can lift the chassis out of the tank and place the filter in the receiver.

[0312] In an example embodiment, the management system can notify the user when the filter needs to be repaired or replaced. For example, the management system can include a timer or counter, which is enabled when the filter is installed on the container. If the management system determines that the filter has been running for more than a threshold time, the management system can send a notification to the user (or other entity). As another example, the management system can enable a timer or counter only when the container is running, the pump is started, or the dielectric fluid passes through the filter (determined by the fluid sensor in the filter). If the management system determines that the filter has been running for more than a threshold time, the management system can send a notification to the user. As another example, the management system can determine the pressure difference of the filter, and if the pressure difference exceeds a threshold pressure, the management system can notify the user to repair or replace the filter. In particular, the filter can include an input pipeline and an output pipeline, and a pressure sensor can be provided on the input pipeline, and a pressure sensor can be provided on the output pipeline. Each pressure sensor can send a pressure reading to the management system. If the pressure difference between the readings of the pressure sensor exceeds the threshold pressure, the management sensor can determine that the filter is clogged. Therefore, the management system can notify the user to repair or replace the filter. As another example, the filter can include a sensor indicating the filter flow rate. The management system can use the flow rate to determine whether the filter needs to be repaired. As yet another example, a management system may use a machine learning model to determine when to replace a filter. The model may receive training data from a central server indicating operational data for filters of a plurality of containers connected to the server.

[0313] In one example embodiment, the liquid collection area and / or the bathtub area may include one or more fluid level sensors. During startup or rapid increase in workload, as the dielectric fluid in the bathtub area evaporates, the fluid level in the liquid collection area decreases. However, because the pump circulates the dielectric fluid, the fluid level in the bathtub area remains unchanged, i.e., the computing device remains submerged. On the contrary, during shutdown or rapid decrease in workload, the fluid level in the liquid collection area may decrease.

[0314] The fluid level sensor can provide data about the fluid level in the sump area and the bath area to the management system. If the fluid level in the sump area decreases after the container is started (or when the container is operating in a steady state), there may be a leak in the tank. Similarly, if the fluid level in the bath area drops at a certain point, there may be a problem with the fluid circulation system, such as a pump failure. In this way, the management system can continuously monitor the fluid level data provided by the fluid level sensor and notify the user if there is an unexpected drop in the fluid level in the sump area or the bath area.

[0315] In an example embodiment, pump can extract fluid from liquid collecting area (or bathtub area), and fluid is provided to the discharge valve connected to tank body.When valve is opened, pump can discharge liquid collecting area (or bathtub area) or provide sample to the user of container.Sample can be provided to laboratory for testing.In an example embodiment, user can use management system to order to discharge tank under container.In response, management system can open discharge valve, and pump can guide fluid from liquid collecting area (or even bathtub area, for example, when being connected to bathtub area) to discharge valve.For example, between bathtub area and liquid collecting area, can be valve connection, and when receiving discharge instruction, valve connection can be connected to bathtub area to liquid collecting area, so that bathtub area discharges dielectric fluid into liquid collecting area, and pump discharges liquid collecting area.In an example embodiment, pump can directly extract dielectric fluid from tank area.

[0316] Robotic system

[0317] In one example embodiment, the container may include a robotic system, such as a gantry robot configured to lift a computing device from a bath area of ​​a tank or from a cassette placed near the tank. The gantry robot may also place the computing device in the bath area or cassette.

[0318] The gantry robot (or robot) may include a series of linear actuators. For example, the robot may include an actuator for moving in each of a plurality of directions (e.g., horizontal and vertical). A management system (or another system) may control how much or how fast each of these actuators moves. In an example embodiment, the actuators may be configured to move on one or more tracks. Actuator-based or track-based systems may lose their accuracy over time (e.g., due to drift or wear and tear). Thus, in this example embodiment, in order to detect the exact relative positioning of the robot, the tank (or its various components) may include one or more calibration areas or markers. For example, one or more key components or key locations of a container that interacts with the robot, such as a casing, a first server rack, a second server rack, or an original positioning, may include a marker that the robot can detect once the robot reaches the positioning of the key component or key location. The marker may inform the robot about the precise position of the robot relative to the key component or location.

[0319] In an example embodiment, the marker can be a physical object, an RFID tag, a color, an alphanumeric code, a QR code, etc. In an example embodiment, the sensor that detects the marker can be a motion sensor, an RFID detector, a camera, etc. In an example embodiment, the camera can determine the distance between the robot and each object and provide feedback to the robot regarding the distance. In an example embodiment, the camera can provide video data to the management system, and based on the video data, the management system can determine the exact position of the robot within the container. In an example embodiment, the management system can determine the position of the robot by, for example, scanning a QR code, counting the parts in the tank, etc. In an example embodiment, the image from the camera can be used to determine the proximity of the robot to the object or determine whether the robot has properly grasped or placed the chassis. In an example embodiment, the management system can use object recognition technology to determine the position of the robot. In an example embodiment, the management system can use artificial intelligence technology to determine the position of the robot. The management system can use object recognition technology or artificial intelligence to calibrate the robot.

[0320] In an exemplary embodiment, the container may include an original position, a case, and two racks. The management system (or another system) may give instructions to the robot to lift the computing device from, for example, the second rack. The robot may move from the original position to the case, and then to the first and second racks. As the robot approaches each of these locations or components, the robot's sensors may detect the associated markings for the location or component. An advantage of the marking system is that the robot can still detect the positioning of the robot relative to a critical component or critical position, even if other components or positions have been removed from the container. This is because the marking is always in the same position relative to each critical component or critical position associated with the mark. For example, even if the first rack is removed from the container, the robot can find the marking of the second rack, calibrate the positioning of the robot relative to the second rack, and remove the computing device from the second rack. Similarly, even if the second rack is slightly moved from its position in the container, the robot can find the marking of the second rack, calibrate the positioning of the robot relative to the second rack, and remove the computing device from the second rack.

[0321] In one example embodiment, the gantry robot can receive instructions to remove or replace various components of the container, such as computing devices, filters, etc. In one example embodiment, the instructions can be provided by a management system (or another system). The management system can provide instructions in response to a determination by the management system (or another system), a user of the container, or a system external to the container. For example, the management system can receive and monitor various data points related to the operation of the computing device, such as voltage levels, temperatures, and other operating characteristics. If the computing device exceeds a threshold determined or predetermined for the computing device, the management system is programmed to give instructions to the robot to replace the computing device.

[0322] As another example, a user of the container may instruct the management system to provide instructions to the robot to remove the computing device. As yet another example, the management system may include an application programming interface (API) for receiving instructions from a system external to the container. For example, the container may communicate with a top-level orchestration and management platform, which may give instructions to the management system (via the API) to remove the computing device from the tank.

[0323] In one example embodiment, the robot can lift a computing device from a tank or a case. In this example embodiment, the computing device can be located in a case that includes a connection plate. The robot can include guide pins and a coupling mechanism that can interface with the connection plate. The robot can also include one or more load cells that measure positive or negative forces or pressures applied to the robot.

[0324] The robot can start at its original location and move toward the tank (or rack including the computing device). At the tank, the robot can detect a marker associated with the tank, which can inform the robot that it is at the tank. The robot can then move a predetermined distance from the marker so that the robot is precisely (or approximately) located above the computing device. Once the robot is on top of the computing device, the robot can quickly descend from its topmost position to a position a few inches away from the computing device (or a position equal to or longer than the length of the guide pin, such as 50% longer than the guide pin). At this point (i.e., a few inches away from the computing device), the robot can more slowly approach the computing device so that the robot's guide pin makes initial contact with the connection plate. Once the robot makes initial contact, the robot will continue to move downward at the same slow speed until the robot presses the connection plate (of the chassis) with a pressure exceeding a threshold. At this point, the robot's connection mechanism can be enabled (e.g., the finger-shaped member can be opened) to connect the robot to the computing device. The connection mechanism can be an armature-based connection mechanism including a plurality of fingers. Once the connection mechanism is closed, the robot can provide feedback to the management system that the connection system is closed. The management system can instruct the robot to lift the computing device. The robot can slowly pull the chassis up a few inches to ensure it has a good grip on the chassis. The robot can then quickly move up to its topmost position. At this point, the robot can move to any location instructed by the management system, such as a chassis or another rack.

[0325] In one example embodiment, the robot can place the computing device in a tank or a case. For example, when the case is held, the robot can move above the slot of the tank (or one of its racks) to lower or place the case into the tank (or case). Once the robot is above the tank, the robot begins to move downward quickly until it reaches a few inches above the first alignment point (or matching point) between the case and the tank (or rack). The design of the case and the tank can determine the distance that the robot should slowly descend above the tank. In particular, the robot can slowly descend one or two inches above the alignment point (the position where the track of the case contacts the groove of the rack). The robot can slowly move toward the tank so that the groove of the rack can move in the track of the case. The management system can receive and monitor data from load cells and other sensors to ensure that the case is not misaligned. For example, excessive force feedback on the load cell may indicate misalignment between the groove and the track. If misalignment is detected, the management system can terminate the lowering operation.

[0326] In an example embodiment, in addition to the grooves and rails, the chassis and the frame may include additional alignment mechanisms. For example, after the initial fit between the chassis and the frame using the grooves and rails, a guide pin mechanism may be provided on the frame and the chassis, which may further align the frame and the chassis. The guide pin mechanism may include a pin on the frame and a mating hole on the chassis. After the initial fit, the robot may move down quickly again until it reaches the second alignment mechanism (or a few inches thereof, for example, two inches larger than the size of the guide pin). Here, the second alignment mechanism may be a guide pin mechanism. The robot moves down slowly so that the pin on the frame may be connected to the mating hole on the chassis. The robot continues to move down slowly until the load cell provides feedback indicating that the chassis has been inserted, for example, the load cell detects a positive pressure. At this point, the coupling mechanism may be disabled, and the robot may move up (slowly moving a few inches to ensure correct placement, and then moving quickly) and return to its original position.

[0327] In one example embodiment, the chassis or rack may include a presence detection pin. When the presence detection pin engages with a corresponding receiver, the management system may receive a signal from the receiver. The signal may indicate that the chassis is correctly placed at the location. In this example embodiment, the robot may deactivate the coupling mechanism only after the receiver provides a signal to the management system.

[0328] During the lifting or lowering operation, the management system can receive and monitor data received from the load cell and other sensors (e.g., motion sensors, tilt sensors, rotation sensors, accelerometers, etc.). This data can ensure that the chassis is not stuck or misaligned, or that the robot has a good grip on the chassis. If the management system determines that the chassis is stuck or misaligned in some way, or that the robot is tilting or rotating (e.g., due to a poor connection between the robot and the chassis), the management system can abort the lifting or lowering operation.

[0329] Fig. 20A -B shows an example robotic system 1100 . Fig. 20A A robotic system 1100 is shown, which may be a gantry robot including a plate 1110. The gantry robot may move within a tank and use the plate 1110 to lift computing equipment. Fig. 20B A plate 1110 is shown, which may include a coupling mechanism 1111 and guide pins 1112. The coupling mechanism 1111 may include a plurality of fingers 1113 that are mechanically coupled to one or more armatures. Once the coupling mechanism is placed in the connection plate of the chassis, the armature may activate and move the fingers to hold the connection plate.

[0330] Fig.21A-B shows an example guide pin mechanism between the chassis 1150 and the rack 1160. In this example embodiment, the rack 1160 may include two guide pins 1161 (for each chassis 1150), and the chassis 1150 may include two mating holes 1151 configured to receive the guide pins 1161. When the robot lowers the chassis 1150 onto the rack 1160, the guide pin mechanism ensures proper electrical connection between the chassis 1150 and the rack 1160.

[0331] In one example embodiment, the robot is a robotic arm. The robotic arm can move on a track disposed on one side of the tank. In one example embodiment, each chassis can be pulled up using a piston, and each chassis can be connected to a channel located on top of the piston. The channel can transport the chassis to the casing, for example, using a track system.

[0332] In one example embodiment, the robot may include a calibration system, which may include a plurality of sensors. The calibration system may determine if the robot is outside of its normal operating range. For example, a tilt sensor may notify the robot if the robot is unbalanced or tilted. As another example, a load cell may provide a signal to the robot if the robot is not moving freely, such as if it has bumped into an object.

[0333] In one example embodiment, the robot may use artificial intelligence or machine learning techniques to provide hot swapping or as a fail-safe mechanism.

[0334] In one example embodiment, the container may include multiple cameras. In this example, one camera may be mounted on the robot and another camera may be mounted on the container wall. The cameras may be mounted in a manner so that a user always has visibility of the moving parts of the container. The container may also include a user interface displayed on a display device such as a monitor. When the robot is lifting or lowering a chassis, the user interface may display a video feed from the camera. This way, if there are any issues with the operation of the robot, the user can take action.

[0335] In one example embodiment, the robotic system may be a vision based system associated with active control. Active control allows for logic to send back a reference point which will determine proximity via a proximity switch. In one example embodiment, the robotic system may be an AI robotic system. In one example embodiment, the robotic system may be an auto-correction system. In one embodiment, the robotic system may be a logic controlled active loop based system which is pre-programmed and can be calibrated based on the distance from its rest state to its active state.

[0336] Absorption / desorption unit

[0337] In one example embodiment, the absorption unit can be a carbon bed based system. The absorption unit can be a round cylindrical drum. Inside the absorption unit there can be an aluminum frame that allows for the inclusion of a copper strip heating element throughout the frame. The height and radius of the absorption unit can be designed based on the size of the container and the volume of fluid in the tank.

[0338] The absorption unit may be sealed and include activated carbon within a frame. The absorption unit may include an inlet and an outlet. In one embodiment, the absorption unit may include a cooling system, for example, cold air may flow through the center of the absorption unit without contacting the carbon. The system allows the carbon to be cooled by convection.

[0339] In an example embodiment, activated carbon allows the absorption or adhesion of dielectric vapor. When the storage tank needs to be balanced (e.g., to generate pressure or vacuum), the management system can be connected to the storage tank by opening a valve by an absorption unit. The management system can enable or start the power supply of the copper heating band element, which can heat the charcoal. Then, the charcoal releases dielectric fluid molecules as steam, which can be returned to the storage tank.

[0340] In an example embodiment, there may be pressure sensors and temperature sensors in the carbon bed to prevent over-pressure or over-temperature conditions.

[0341] In an example embodiment, the absorption unit may include a control loop bypass for pressure relief (or pressurization) in an emergency. This is a safety feature of the container. The absorption unit has a valve that disconnects the absorption unit from the tank. If the valve fails, an overpressure situation may occur. For example, if the outlet of the absorption unit is blocked, the management system can open the bypass valve. If the bypass valve is opened, dielectric vapor can enter the atmosphere, thereby preventing the absorption unit from exploding. If there is a vacuum condition in the tank, the valve can be opened to allow air to enter the tank, thereby preventing the tank from collapsing.

[0342] In one example embodiment, the container may include multiple safety bypass valves. For example, during startup operations, the computing equipment may generate excess steam. The valve that allows the steam to leave the tank to the absorption unit may fail. As a result, there may be an overpressure condition in the tank that the bellows cannot handle. The emergency bypass valve may be opened to release some of the steam to the atmosphere.

[0343] As another example, during startup operation, excess steam may enter the absorption unit. This may create an overpressure condition in the absorption unit. As such, a bypass valve of the absorption unit may be opened to release the steam to the atmosphere.

[0344] In one example embodiment, the management system can receive and monitor data from the absorption unit related to the electrical power of the absorption unit in addition to pressure and temperature. The management system ensures that current flows through the absorption unit. If there is an overcurrent problem or an overvoltage condition exists, the management system can shut down the absorption unit.

[0345] Self-alignment of the chassis

[0346] In one example embodiment, the chassis may include self-alignment features. The self-alignment features may include a plate that may move (i.e., float) relative to the chassis. The plate may have one or more input ports or output ports (or connectors). The chassis (and plate) may include mating holes that may receive guide pins to align the plate to receive the ports. In one example embodiment, the ports may be Connector.

[0347] In one example embodiment, the self-alignment feature may include several alignment mechanisms. For example, as a first self-alignment mechanism, the rack and chassis may have grooves and guide rails. As a second self-alignment mechanism, the rack may include a pin that is tapered and rounded at the end. The pin may enter a locking hole in the chassis. When the pin is inserted into the mating hole (or locking hole), it will perform a final precise alignment between the connector on the chassis and the mating part (i.e., the backplane) on the rack. When the connector is ready to mate with its mate, the alignment pin has perfectly matched the relative orientation of the floating mating piece to the connector it is about to be inserted into.

[0348] In one example embodiment, the connectors of the chassis and the rack may include their own alignment mechanisms, for example, pins may be part of the connectors.

[0349] In one example embodiment, the connector may include a multi-stage mechanism for self-alignment, including a coarse outer alignment detent followed by a finer inner alignment detent.

[0350] In one example embodiment, the board may be on a backplane of a rack.

[0351] Fig. 22 Example connectors with self-aligning features are shown. These connectors can include guide pins and other guide features to ensure proper connection between mating parts.

[0352] Immersion Cooling Process

[0353] 1. A method comprising:

[0354] A computer component is at least partially immersed in a thermally conductive, condensable dielectric fluid, wherein:

[0355] The computer components are mounted in a chassis, the chassis including a backplane for receiving power from a rack; and

[0356] The computer component is configured to dissipate heat in a dielectric fluid when the computer component is operating;

[0357] condensing a gas phase of the dielectric fluid into a liquid phase of the dielectric fluid using a condenser;

[0358] Wherein, the frame is within a tank including a pressure controller for reducing or increasing the internal pressure of the tank.

[0359] 2. The method of embodiment 1, wherein the storage tank has a calculated power density of at least 300 W of calculated power dispersed per square foot of space.

[0360] 3. The method according to embodiment 1 further includes removing the chassis from the rack using a robot, wherein the robot is located in the storage box.

[0361] 4. The method of embodiment 3 further comprises delivering the chassis to an airlock using the robot, wherein the airlock is configured to allow access to the interior of the tank without significantly disrupting the pressure within the tank.

[0362] 5. The method according to embodiment 4 further comprising:

[0363] opening an inner door of said airlock;

[0364] placing the chassis in the airlock;

[0365] closing the inner door of the airlock;

[0366] equalizing the airlock pressure to atmospheric pressure; and

[0367] Open the outer door of the airlock.

[0368] 6. The method of embodiment 3 further comprises storing the chassis in a cassette using the robot.

[0369] 7. The method of embodiment 6, wherein the casing is located on a platform including a support member, a rotating member, and a track.

[0370] 8. The method of claim 3, wherein the robot is a gantry robot configured to remove, replace, or assemble the chassis.

[0371] 9. A method according to embodiment 8, wherein the gantry robot is configured to move on a horizontal plane and descend vertically.

[0372] 10. The apparatus of embodiment 9, wherein the robot is configured to remove, replace, or install components of a power distribution system.

[0373] 11. The method of embodiment 10, wherein the robot comprises a gripping tool for gripping the chassis.

[0374] 12. The method of embodiment 1, wherein the tank is mounted within a super structure comprising a plurality of tanks.

[0375] 13. The method of embodiment 1 further comprising removing contaminants from the dielectric fluid.

[0376] 14. The method of embodiment 1 further comprising removing gaseous pollutants.

[0377] 15. The method of embodiment 1 further comprising providing power, network connectivity, and process fluid to the storage tank.

[0378] 16. The method of embodiment 1, wherein the tank comprises an opening on the top and a removable cover.

[0379] 17. The method of embodiment 1, wherein the tank comprises an internal volume between about 100 cubic feet and about 300 cubic feet.

[0380] 18. The method of embodiment 1, wherein the chassis does not include a heat sink and a fan.

[0381] 19. The method of embodiment 1, wherein the chassis includes a blade server, a processor, a power supply, or an interface card.

[0382] 20. The apparatus of embodiment 19, wherein the backplane is electrically connected to an interface card, the interface card being a Cat6A or Cat7 compatible RJ45 interface for connecting to a 1G or 10G Ethernet interface. Container Design and Configuration for Immersion Cooling

[0383] 1. A device comprising:

[0384] a tank configured to contain a thermally conductive, condensable dielectric fluid;

[0385] a pressure controller for reducing or increasing the internal pressure of the tank;

[0386] a frame at least partially immersed in the dielectric fluid;

[0387] a condenser for condensing the gas phase of the dielectric fluid; and

[0388] A robot is configured to move the chassis within the rack.

[0389] 2. The apparatus of embodiment 1, wherein the apparatus comprises a modular sled comprising a plurality of fork lift tubes.

[0390] 3. The apparatus of embodiment 1, wherein the storage tank has a calculated power density of at least 300 W of calculated power dispersed per square foot of space.

[0391] 4. The device of embodiment 1, wherein the exterior of the device includes a power input and a communication input.

[0392] 5. The device according to claim 4, wherein:

[0393] The power input and the communication input are electrically connected to the box body; and

[0394] The box distributes the power input and communication input to the racks using a plurality of wires.

[0395] 6. The apparatus of embodiment 5, wherein the rack comprises a backplane receiver configured to distribute power signals and communication signals to the chassis.

[0396] 7. The apparatus of embodiment 6, wherein the chassis includes a backplane configured to: receive power signals and communication signals from a backplane receiver of the rack; and

[0397] Power signals and communication signals are distributed to computer components within the chassis.

[0398] 8. The device of embodiment 5, wherein the plurality of wires does not include a plastic insulator.

[0399] 9. The apparatus of embodiment 5, wherein the rack comprises a transformer.

[0400] 10. The device of embodiment 1, wherein the device is stackable.

[0401] 11. The device of embodiment 1, wherein the device includes a receiver for storing replacement parts.

[0402] 12. The apparatus of embodiment 11, wherein the robot is configured to remove the chassis from the rack and place the chassis in the cassette.

[0403] 13. The apparatus of embodiment 12, wherein the casing is located on a platform, the platform comprising a rotating member, a supporting member, and a track.

[0404] 14. The device of embodiment 13, wherein the platform is configured to guide the casing to the exterior of the device.

[0405] 15. The device of embodiment 1, wherein the device comprises a desiccant configured to remove water vapor contaminants from the device.

[0406] 16. The apparatus of embodiment 1, further comprising:

[0407] Liquid collection area;

[0408] Pumps; and

[0409] Filters;

[0410] Wherein the pump is configured to remove the dielectric fluid from the sump area and pass the dielectric fluid through the filter before delivering the dielectric fluid to the bath portion of the tank.

[0411] 17. The device of embodiment 1 wherein the dielectric fluid has a boiling point in the range of 20°C to 100°C.

[0412] 18. The device of embodiment 1, wherein the dielectric fluid comprises a chemical having a chemical formula of (CF3)2CFCF2OCH3, C4F9OCH3, CF3CF2CF2CF2OCH3, a hydrofluoroether, or methoxy-nonafluorobutane.

[0413] 19. The device of embodiment 1 further comprising a lock that prevents the device from operating if either a cover or a door of the device is not secured.

[0414] 20. The apparatus of embodiment 1 further comprising a controller configured to de-energize the device in the event of unauthorized access to the cover or door.

[0415] Robotics and Automation for Immersion Cooling

[0416] 1. A device comprising:

[0417] a tank configured to contain a thermally conductive, condensable dielectric fluid;

[0418] a pressure controller for reducing or increasing the internal pressure of the tank;

[0419] a computer component at least partially immersed in the dielectric fluid;

[0420] a condenser for condensing the gas phase of the dielectric fluid; and

[0421] A robot is configured to pick up the computer component.

[0422] 2. The device according to embodiment 1 also includes an airlock.

[0423] 3. The apparatus of embodiment 2, wherein the airlock comprises an inner door and an outer door.

[0424] 4. The apparatus of embodiment 3, wherein the airlock is configured to receive an inert gas to purge a gas phase of the dielectric fluid before the outer door is opened.

[0425] 5. The apparatus of embodiment 3, wherein the robot is located outside the tank.

[0426] 6. The apparatus of embodiment 3, wherein the robot is located within the tank.

[0427] 7. The apparatus of embodiment 6, wherein the robot is configured to remove computer components from a rack and deliver the computer components to the airlock.

[0428] 8. The apparatus of embodiment 7, wherein the robot is further configured to:

[0429] opening an inner door of said airlock;

[0430] placing the computer component in the airlock;

[0431] closing the inner door of the airlock;

[0432] equalizing the airlock pressure to atmospheric pressure; and

[0433] Open the outer door of the airlock.

[0434] 9. The apparatus of embodiment 8 further comprises a second robot located outside of the tank.

[0435] 10. The apparatus of embodiment 9, wherein the second robot is configured to remove the computer component from the airlock when the exterior door is open.

[0436] 11. The apparatus of embodiment 9, wherein the second robot is configured to place the computer component in a storage slot.

[0437] 12. The apparatus of embodiment 9, wherein the airlock is configured to balance the pressure of the airlock after the outer door is closed.

[0438] 13. The apparatus of embodiment 1, wherein the apparatus is configured to receive instructions from a server external to the apparatus.

[0439] 14. The apparatus of embodiment 1, wherein the computer component is located within a chassis displaying an asset tag.

[0440] 15. The apparatus of embodiment 14, wherein the robot is configured to scan the asset tag and forward the asset tag to a management system.

[0441] 16. The apparatus of embodiment 1, wherein the robot is a gantry robot configured to remove, replace, or assemble the computer components.

[0442] 17. The apparatus of embodiment 8, wherein the gantry robot is configured to move horizontally and vertically.

[0443] 18. The apparatus of embodiment 1, wherein the robot is configured to remove, replace, or install components of an electrical power distribution system.

[0444] 19. The apparatus of embodiment 10, wherein the component of the power distribution system is a transformer or a power supply.

[0445] 20. The apparatus of embodiment 1, wherein the robot comprises a gripping tool for gripping the computer component.

[0446] Ballast blocks for immersion cooling

[0447] 1. A device comprising:

[0448] Tank, including:

[0449] a bath portion for containing a thermally conductive, condensable dielectric fluid and computer components; and

[0450] a shelf portion configured to hold at least one ballast mass;

[0451] a pressure controller for reducing or increasing the internal pressure of the tank;

[0452] a condenser for condensing the gas phase of the dielectric fluid; and

[0453] A robot is configured to pick up the computer component.

[0454] 2. The apparatus of embodiment 1, wherein a bottom point of the bath portion has a lower height than a height of the shelf portion.

[0455] 3. The apparatus of embodiment 1, wherein the bath portion is configured to at least partially immerse the computer component in the dielectric fluid.

[0456] 4. The apparatus of embodiment 3, wherein the computer component is a blade server, a processor, a power supply, or a transformer.

[0457] 5. The apparatus of embodiment 1, wherein the level of the dielectric fluid is high enough to cover at least a portion of the shelf portion.

[0458] 6. The apparatus of embodiment 1, wherein the shelf portion is adjacent to the condenser.

[0459] 7. The apparatus of embodiment 6, wherein the shelf portion is configured to receive condensed dielectric fluid from the condenser.

[0460] 8. The apparatus of embodiment 1 wherein the ballast mass is configured to occupy a volume above a shelf of the tank to transfer the dielectric fluid from the shelf to an area above a bath portion.

[0461] 9. The apparatus of embodiment 1 wherein the ballast mass includes a plurality of lifter feet for allowing dielectric fluid to flow beneath the ballast mass.

[0462] 10. The device of embodiment 1 wherein the ballast mass is insoluble in the dielectric fluid.

[0463] 11. The device of embodiment 1, wherein the ballast mass is made of metal, rubber, silicone, or a polymer.

[0464] 12. The device of embodiment 1 wherein the ballast mass is denser than the dielectric fluid.

[0465] 13. The device of embodiment 1 wherein the ballast mass has a handle, cutout, or plate for removing or replacing the ballast mass.

[0466] 14. The apparatus of embodiment 13, wherein the robot is configured to lift the ballast mass using the handle, cutting portion, or plate.

[0467] 15. The device of embodiment 1, wherein the ballast mass is configured to interlock with another ballast mass from a top side or a bottom side of the ballast mass.

[0468] 16. The device of embodiment 15, wherein the interlock prevents the other ballast mass from sliding.

[0469] 17. The apparatus of embodiment 15, wherein the further ballast mass is configured on a top side or a bottom side of the ballast mass.

[0470] 18. The apparatus of embodiment 15, wherein the ballast mass includes a recessed portion on a top side of the ballast mass such that a lift foot of another ballast mass is configured to lock into one of the recessed portions of the ballast mass.

[0471] 19. The apparatus of embodiment 1 wherein the ballast mass is configured to span at least 40% of the entire length of the shelf portion.

[0472] 20. The apparatus of embodiment 1 wherein the ballast mass has outer dimensions of approximately 2 feet in length, approximately 8 inches in width, and approximately 1 inch in height.

[0473] Server enclosures for immersion cooling

[0474] 1. A device comprising:

[0475] a tank configured to contain a thermally conductive, condensable dielectric fluid;

[0476] a pressure controller for reducing or increasing the internal pressure of the tank;

[0477] a chassis at least partially immersed in the dielectric fluid;

[0478] a condenser for condensing the gas phase of the dielectric fluid; and

[0479] A robot is configured to pick up the chassis.

[0480] 2. The apparatus of embodiment 1, wherein the chassis does not require a heat sink or a fan.

[0481] 3. The apparatus of embodiment 1, wherein the chassis comprises a blade server.

[0482] 4. The apparatus of embodiment 1, wherein the chassis includes a processor, a power supply, or an interface card.

[0483] 5. The apparatus of embodiment 19, wherein the interface card is a Cat6A or Cat7 compatible RJ45 interface for connecting to a 1G or 10G Ethernet interface.

[0484] 6. The apparatus of embodiment 1, wherein the chassis is removably attached to a rack.

[0485] 7. The apparatus of embodiment 6, wherein the chassis includes a backplane to provide a slot-type interface between the chassis and the rack.

[0486] 8. The apparatus of embodiment 7, wherein the backplane is configured to distribute power and signals received from the rack within the chassis.

[0487] 9. The apparatus of embodiment 8, wherein the backplane is configured to transmit power and data to the blade servers via cables.

[0488] 10. The apparatus of embodiment 1, wherein the chassis is a substantially rectangular box including a rear wall and two side walls, wherein the rear wall has a plurality of openings to facilitate the circulation of dielectric fluid within the chassis.

[0489] 11. The apparatus of embodiment 10, wherein the chassis includes guide rails on each of two side walls.

[0490] 12. The apparatus of embodiment 1, wherein the chassis includes mounting interfaces for retaining computer components.

[0491] 13. The apparatus of embodiment 1, wherein the chassis includes a panel and the robot is configured to lift the chassis using the panel.

[0492] 14. The apparatus of embodiment 1, wherein the chassis comprises a microcontroller.

[0493] 15. The apparatus of embodiment 14, wherein the microcontroller is configured to:

[0494] receiving sensor data from a sensor mounted on the chassis, the sensor data indicating whether the chassis is properly positioned in a rack; and

[0495] The sensor data is transmitted to a management system.

[0496] 16. The apparatus of embodiment 14, wherein the microcontroller is configured to:

[0497] receiving a power signal from a management system; and

[0498] The power signal is transmitted to a switch configured to shut off power within the chassis.

[0499] 17. The apparatus of embodiment 14, wherein the microcontroller is configured to:

[0500] receiving operational data from a computer component mounted within the chassis; and

[0501] The operating data is transmitted to the management system.

[0502] 18. The apparatus of embodiment 14, wherein the microcontroller is configured to control electrical and communication facilities of a blade server.

[0503] 19. The apparatus of embodiment 1, wherein the chassis includes an RFID tag.

[0504] 20. The apparatus of embodiment 19, wherein the robot is configured to scan the RFID tag and transmit a signal to a management system.

[0505] Steam management for immersion cooling using bellows

[0506] 1. A device comprising:

[0507] a tank configured to contain a thermally conductive, condensable dielectric fluid and computer components;

[0508] a pressure controller for reducing or increasing the internal pressure of the tank;

[0509] a vapor management system for condensing a vapor phase of the dielectric fluid; and

[0510] A robot is configured to pick up the computer component.

[0511] 2. The apparatus of embodiment 1, wherein the vapor management system comprises a condensation structure within the storage tank.

[0512] 3. The device of embodiment 2, wherein the condensing structure comprises a heat pipe, a coil, or a radiator fin.

[0513] 4. The apparatus of embodiment 2, wherein the condensation structure is configured to be coupled to a cooling liquid source such that the cooling liquid passes through the condensation structure.

[0514] 5. The apparatus of embodiment 2, wherein the apparatus is configured to refrigerate cool the liquid using evaporative cooling or a dry cooling tower.

[0515] 6. The apparatus of embodiment 2, wherein the steam management system comprises an inflow conduit and an outflow conduit.

[0516] 7. The apparatus of embodiment 6, wherein the inflow conduit is configured to receive cooling liquid from a refrigerated cooling liquid source and direct the cooling liquid through a condensing structure.

[0517] 8. The apparatus of embodiment 6, wherein the outflow conduit is configured to receive cooling liquid from the condensing structure and return the cooling liquid to a refrigerated cooling liquid source.

[0518] 9. The apparatus of embodiment 1, wherein the vapor management system comprises a storage unit for storing a dielectric fluid.

[0519] 10. The apparatus of embodiment 9, wherein the vapor management system is configured to direct dielectric fluid from the storage unit into the tank.

[0520] 11. The apparatus of embodiment 1, wherein the vapor management system comprises a vapor storage unit for storing vapor of the dielectric fluid.

[0521] 12. The device of embodiment 11, wherein the vapor storage unit is a bellows.

[0522] 13. The device of embodiment 12, wherein the bellows is configured to expand or contract to maintain the internal pressure of the tank.

[0523] 14. The device of embodiment 12, wherein the bellows comprises one or more reservoir bags.

[0524] 15. The device of embodiment 11, wherein the vapor storage unit comprises a valve for allowing air to enter the vapor management system to reduce the temperature of the vapor of the dielectric fluid.

[0525] 16. The apparatus of embodiment 15, wherein the vapor storage unit is operably connected to the carbon bed to separate the vapor of the dielectric fluid from the air.

[0526] 17. The apparatus of embodiment 16, wherein the carbon bed comprises a desorption heater configured to heat the carbon bed to increase the temperature of the carbon bed.

[0527] 18. The apparatus of embodiment 1 wherein the vapor management system comprises a filter.

[0528] 19. The device of embodiment 17, wherein the filter is configured to remove air and water vapor.

[0529] 20. The apparatus of embodiment 1, wherein the steam management system:

[0530] includes an inert gas storage unit; and

[0531] The inert gas storage unit is configured to introduce inert gas into the tank from the inert gas storage unit during a startup operation or a shutdown operation.

Claims

1. A method comprising: A computer component is at least partially immersed in a thermally conductive, condensable dielectric fluid, wherein: The computer components are mounted in a chassis, the chassis including a backplane for receiving power from a rack; and The computer component is configured to dissipate heat in a dielectric fluid when the computer component is operating; condensing a gas phase of the dielectric fluid into a liquid phase of the dielectric fluid using a condenser; Wherein, the frame is within a storage tank, and the storage tank comprises a pressure controller for reducing or increasing the internal pressure of the storage tank.

2. A device comprising: a tank configured to contain a thermally conductive, condensable dielectric fluid; a pressure controller for reducing or increasing the internal pressure of the tank; a frame at least partially immersed in the dielectric fluid; a condenser for condensing the gas phase of the dielectric fluid; and A robot is configured to move the chassis within the rack.

3. A device comprising: a tank configured to contain a thermally conductive, condensable dielectric fluid; a pressure controller for reducing or increasing the internal pressure of the tank; a computer component at least partially immersed in the dielectric fluid; a condenser for condensing the gas phase of the dielectric fluid; and A robot is configured to pick up the computer component.

4. A device comprising: Tank, including: a bath portion for containing a thermally conductive, condensable dielectric fluid and computer components; and a shelf portion configured to hold at least one ballast mass; a pressure controller for reducing or increasing the internal pressure of the tank; a condenser for condensing the gas phase of the dielectric fluid; and A robot is configured to pick up the computer component.

5. A device comprising: a tank configured to contain a thermally conductive, condensable dielectric fluid; a pressure controller for reducing or increasing the internal pressure of the tank; a chassis at least partially immersed in the dielectric fluid; a condenser for condensing the gas phase of the dielectric fluid; and A robot is configured to pick up the chassis.

6. A device comprising: a tank configured to contain a thermally conductive, condensable dielectric fluid and computer components; a pressure controller for reducing or increasing the internal pressure of the tank; a vapor management system for condensing a vapor phase of the dielectric fluid; and A robot is configured to pick up the computer component.

Images