Two-phase immersion cooling system

Abstract

A two-phase immersion cooling system is provided that includes a tank having a tank cavity. The tank is configured to contain a coolant liquid and a computing system submerged in the coolant liquid. The tank further includes bellows coupled to a top plate of the tank, and one or more pressure relief valves are disposed on the bellows and the tank. A pair of power racks are coupled to the tank by at least one busbar, and a pair of grounding assemblies are disposed on outer walls of the tank. A multi-purpose sleeve for securing the gasket a server chassis is also provided.

Description

TWO-PHASE IMMERSION COOLING SYSTEMA. CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of and priority to U.S. Provisional Application No. 63 / 516,996, filed August 1, 2023, and U.S. Provisional Application No. 63 / 519,223, filed August 11, 2023, which applications are hereby incorporated by reference in their entireties.B. BACKGROUND

[0002] Section 1 : Two-phase immersion cooling is a liquid cooling technique that utilizes a liquid- to-gas phase transition to remove heat from a heat-generating component, such as a processor or, more generally, an electronic device of a computing system. This is typically accomplished using a tank to submerge the heat-generating component in a pool of coolant liquid where the coolant liquid is a dielectric with a relatively low boiling point (e.g., 50°C). During operation, the coolant liquid in direct contact with the heat-generating component vaporizes (i.e., boils), producing coolant vapor that rises upwards within the tank. The coolant vapor thereafter transfers heat to a heat exchanger (e.g., a condenser) disposed above the coolant liquid, thus causing the coolant vapor to condense back to a liquid and thereafter fall into the pool of coolant liquid below. A two-phase immersion cooling system is often used to cool high density computing systems, such as an array of servers in a data center.

[0003] Section 2: One challenge with operating a data center is keeping the data center’ s servers and other computing components cool enough to operate quickly and efficiently. Dissipating the heat generated by the servers becomes more challenging as the number of servers grows and the operating temperatures of the servers increase. Unfortunately, air cooling is usually not sufficient to cool the large number of servers in a modern data center.

[0004] Fortunately, it is feasible to cool servers by immersing or submerging them in a thermally conductive dielectric liquid coolant. This type of cooling, called immersion cooling, offers several advantages over air cooling. First, it works without cooling fins or fans on the servers or other computing components being cooled, reducing the servers’ size and weight. Second, it works without a refrigerated air conditioning system, reducing data center power consumption. Third, the liquid coolant has a higher heat capacity than air, so a given volume of liquid coolant can dissipate more heat than the same volume of air.

[0005] There are two types of immersion cooling: single-phase immersion cooling and two-phase immersion cooling. In single-phase immersion cooling, the servers are immersed or submerged inliquid coolant that circulates between a heat exchanger and an immersion tank that contains the servers and immersion fluid. As it circulates, the liquid coolant moves heat generated by the servers to a cold-water circuit or other heat sink thermally coupled to the heat exchanger. The coolant remains in the liquid phase as it circulates, hence the name “single-phase” immersion cooling.[0006| In two-phase immersion cooling, the servers are submerged in liquid coolant with a relatively low boiling point (e.g., at or about 50 °C). The liquid coolant absorbs heat generated by the servers; this causes the liquid coolant to evaporate. The coolant vapor rises from the surface of the liquid coolant, moving heat away from the servers immersed in the liquid coolant. The coolant vapor is cooled by a heat exchanger, such as a condenser coil, and returns to the liquid phase in the immersion tank holding the servers and the liquid coolant.C. SUMMARY

[0007] Section 1: The present disclosure is directed to a two-phase immersion cooling provides an energy efficient approach to cool a computing system, such as a high density computing system. However, recognized conventional two-phase immersion cooling systems (also referred to herein as an “immersion cooling system”) have only been demonstrated for relatively lower power computing systems that have a power density less than 250 kilowatts (kW) per enclosure. For relatively higher power computing systems (i.e. , computing systems having a power density greater than 250 kW per enclosure), the operating conditions of the immersion cooling system can be appreciably more challenging to manage.

[0008] For instance, higher power computing systems generate appreciably higher heat loads. This, in turn, can lead to appreciably higher coolant vapor generation rates during operation and thus large pressure fluctuations within the tank. As a result, the risk of overpressure -related failures (i.e., when the tank pressure is appreciably greater than the ambient pressure) and negative pressure-related failures (i.e., when the tank pressure is appreciably less than the ambient pressure) is appreciably higher for higher power computing systems.

[0009] The concentration of contaminants in the coolant liquid of a conventional immersion cooling system also typically increases over time. The contaminants typically originate from conventional electronic components submerged in the coolant liquid, such as printed circuit boards (PCBs), cables, wires, and / or the like. These contaminants can accumulate onto various surfaces of the computing system that dissipate heat (e.g., electrical interconnects, a ball grid array) resulting in performance degradation and, in some instances, failure of an electronic component in the computing system. For higher power computing systems, contamination is made worse by thepresence of computing components that generate higher heat loads and / or a larger number of heatgenerating components.

[0010] Addressing the foregoing conditions is made even more challenging by the general desire for immersion cooling systems to remain relatively compact in size. A relatively compact immersion cooling system can facilitate greater space utilization, which, in turn, allows for higher computing densities within a facility (e.g., by allowing a greater number of immersion cooling systems to be deployed in a given space). Additionally, conventional coolant liquids for two-phase immersion cooling systems are expensive. Thus, a more compact immersion cooling system that consumes relatively less coolant liquid over time is often preferable.

[0011] The present disclosure is thus directed to various implementations of a two-phase immersion cooling system that includes multiple improvements to facilitate support of higher power computing systems. For example, the two-phase immersion cooling systems disclosed herein may support computing systems having a power density greater than 250 kW or, more preferably, greater than 500 kW. The improvements include, but are not limited to, greater ease of installing and / or removing components of a computing system (e.g., an assembled unit) to / from the tank of the cooling system, more efficient utilization of coolant liquid by using an external power source to provide electrical power to the computing system, a condenser capable of supporting heat loads greater than or equal to 600 kW, improved regulation of tank pressure using bellows and pressure relief valves, and improved regulation of contaminants in the coolant liquid using an onboard filtration system and desiccant assembly.

[0012] Section 2: Embodiments of the present technology include a gasket for a server chassis that holds a server and is installed in an immersion tank filled with liquid coolant. The gasket is formed of an elastically deformable material and has at least one fastener for securing the gasket to the server chassis. The gasket prevents the server chassis from rubbing against a metal surface of the immersion tank as the server chassis and the gasket are inserted into the immersion tank. The gasket can have a chamfered corner. The gasket can be elastically compressed and may be made of plastic, rubber, silicone, plastic, and / or polytetrafluoroethylene. It may include at least one compressible spring feature configured to be compressed between the server chassis and an immersion tank. The gasket can be made of a sponge-like material. The gasket can define a hollow, compressible lumen having a vent or hole to allow fluid flow into and out of the hollow, compressible lumen. The vent or hole can be disposed at an end of the gasket.

[0013] It should be appreciated that all combinations of the foregoing concepts and additional concepts discussed in greater detail below (provided such concepts are not mutually inconsistent) are contemplated as being part of the inventive subject matter disclosed herein. In particular, allcombinations of claimed subject matter appearing at the end of this disclosure are contemplated as being part of the subject matter disclosed herein. It should also be appreciated that terminology explicitly employed herein that also may appear in any disclosure incorporated by reference should be accorded a meaning most consistent with the particular concepts disclosed herein.D. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The skilled artisan will understand that the drawings primarily are for illustrative purposes and are not intended to limit the scope of the inventive subject matter described herein. The drawings are not necessarily to scale; in some instances, various aspects of the inventive subject matter disclosed herein may be shown exaggerated or enlarged in the drawings to facilitate an understanding of different features. In the drawings, like reference characters generally refer to like features (e.g., functionally similar and / or structurally similar elements).

[0015] FIG. 1.1 A shows a front view of an example two-phase immersion cooling system.

[0016] FIG.1. IB shows a top, front, right-side perspective view of the two-phase immersion cooling system of FIG. 1.1 A. The power racks are not shown.

[0017] FIG. 1.2A shows a top, front, right-side perspective view of a tank in the two-phase immersion cooling system of FIG. 1.1 A. The tank is shown with a pair of busbars A.

[0018] FIG. 1.2B shows a top, rear, left-side perspective view of the tank of FIG. 1.2A.

[0019] FIG. 1.2C shows a top view of the tank of FIG. 1.2A.

[0020] FIG. 1.2D shows a front view of the tank of FIG. 1.2A.

[0021] FIG. 1.2E shows a rear view of the tank of FIG. 1.2A.

[0022] FIG. 1.2F shows a right-side view of the tank of FIG. 1.2A.

[0023] FIG. 1.2G shows a left-side view of the tank of FIG. 1.2A.

[0024] FIG. 1.2H shows a cross-sectional perspective view of the tank of FIG. 1 ,2A corresponding to the plane A-A of FIG. 1.2C.

[0025] FIG. 1.21 shows a cross-sectional elevation view of the tank of FIG. 1.2H.

[0026] FIG. 1.2J shows a cross-sectional elevation view of the tank of FIG. 1.2A corresponding to the plane B-B of FIG. 1.2C.

[0027] FIG. 1.3 shows a cross-sectional perspective view of the tank of FIG. 1.2A with multiple assembled units inserted into a tank cavity of the tank.

[0028] FIG. 1.4A shows a cross-sectional perspective view of a tank identical to the tank of FIG. 1 ,2A corresponding to the plane A-A of FIG. 1.2C. The tank is shown with a pair of busbars B .

[0029] FIG. 1.4B shows a cross-sectional elevation view of the tank of FIG. 1.4A.

[0030] FIG. 1.5A shows a photograph of an example two-phase immersion cooling system that includes a tank with a lid in an open position.

[0031] FIG. 1.5B shows a photograph of multiple assembled units inserted into a tank cavity of the tank of FIG. 1.5A.

[0032] FIG. 1.5C shows a photograph of another view of the tank of FIG. 1.5A with the lid in the open position.

[0033] FIG. 1.5D shows a photograph of example control electronics for the two-phase immersion cooling system of FIG. 1.5A.

[0034] FIG. 1.5E shows a photograph of an example filtration system for the two-phase immersion cooling system of FIG. 1.5A.

[0035] FIG. 1.6A shows a top, front, left-side perspective view of the tank of FIG. 1.2A where a lid of the tank is removed to show the tank cavity and, in particular, guide rails disposed within the tank cavity.

[0036] FIG. 1 ,6B shows a magnified top, front, left-side perspective view of the tank of FIG. 1.6A.

[0037] FIG. 1.6C shows a top view of the tank of FIG. 1.6A.

[0038] FIG. 1.7 A shows a cross-sectional perspective view of the tank of FIG. 1.6A with multiple assembled units inserted into the tank cavity via the guide rails.

[0039] FIG. 1.7B shows a cutaway view of the tank of FIG. 1.6A with an assembled unit mounted to the busbar A.

[0040] FIG. 1.7C shows a magnified cutaway view of FIG. 1.7B. Electrical pins and guide pins of the assembled unit are shown inserted into corresponding electrical sockets of the busbar A and guide sockets of a tray.

[0041] FIG. 1.7D shows a cross-sectional view of the assembled unit and the busbar A of FIG. 1.7C.

[0042] FIG. 1.8 shows a bottom perspective view of a portion of the assembled unit of FIG. 1.7A.

[0043] FIG. 1.9 shows a cutaway view of the tank of FIG. 1.6A with an alignment fixture inserted into the tank cavity.

[0044] FIG. 1.10 shows a top perspective view of the tank of FIG. 1.2A with the lid in the open position. The tank is shown with a cable routing bracket and multiple assembled units with cables routed to a passthrough module via the cable routing bracket.

[0045] FIG. 1.11 shows a perspective view of an example passthrough module.

[0046] FIG. 1.12A shows a contour map of deformation across a portion of the tank of FIG. 1.2A when the tank cavity is at a gauge pressure of -100 mbar. The contour map is obtained using Finite Element Analysis (FEA).

[0047] FIG. 1.12B shows another contour map for the portion of the tank of FIG. 1.12A.

[0048] FIG. 1.13 shows a right-side view of the tank of FIG. 1.2A where the busbar A and the busbar B are inserted through respective busbar feedthroughs to illustrate their placement on the tank. In practice, the tank only includes either busbar A or busbar B .[0049| FIG. 1.14A shows a perspective view of the busbar B of FIG. 1.13.

[0050] FIG. 1.14B shows a cross-sectional perspective view of the first and second base bars of the busbar B of FIG. 1.14A corresponding to the plane A-A of FIG. 1.14A. The busbar B is shown inserted through a tray on the tank of FIG. 1.2A.

[0051] FIG. 1.15A shows a perspective view of the busbar A of FIG. 1.13

[0052] FIG. 1.15B shows a cross-sectional perspective view of the multilayer base bar of the busbar A of FIG. 1.15A corresponding to the plane A-A of FIG. 1.15 A. The busbar A is shown inserted through a tray on the tank of FIG. 1.2A.

[0053] FIG. 1.15C shows a cross-sectional perspective view of the multilayer base bar of the busbar A of FIG. 1.15A corresponding to the plane B-B of FIG. 1.15A.

[0054] FIG. 1.16A shows a perspective view of the two-phase immersion cooling system of FIG. 1.1 A where the busbar A is shown coupled to the tank and a power rack via an external multilayer power bar and a feedthrough assembly.

[0055] FIG. 1.16B shows a top, front, right-side perspective view of the feedthrough assembly and the external multilayer power bar of FIG. 1.16A. The power rack is not shown.

[0056] FIG. 1.16C shows another top, front, right-side perspective view of the feedthrough assembly and the external multilayer power bar of FIG. 1.16A.

[0057] FIG. 1.16D shows a cross-sectional perspective view of the busbar A, the feedthrough assembly, and the external multilayer power bar of FIG. 1.16A.

[0058] FIG. 1.16E shows a perspective view of the busbar A, the external multilayer power bar, and the feedthrough assembly. The tank and the power rack are not shown.

[0059] FIG. 1.17A shows a perspective view of the external multilayer power bar and the feed- through assembly of FIG. 1.16A.

[0060] FIG. 1.17B shows a perspective view of the external multilayer power bar.

[0061] FIG. 1.17C shows a cross-sectional perspective view of the external multilayer power bar.

[0062] FIG. 1.18 shows a diagram of the various electrical connections between the power racks and the components of the tank in the two-phase immersion cooling system of FIG. 1.1 A.

[0063] FIG. 1.19 shows an example circuit diagram for the two-phase immersion cooling system of FIG. 1.1 A.

[0064] FIG. 1.20A shows a cross-sectional view of the tank of FIG. 1.4A corresponding to the plane A- A of FIG. 1.4B.

[0065] FIG. 1.20B shows a cross-sectional perspective view of a condenser in the tank of FIG. 20A corresponding to the plane A- A of FIG. 1.20A.

[0066] FIG. 1.20C shows a cross-sectional view of the condenser of FIG. 1.20A corresponding to the plane A- A of FIG. 1.20A.

[0067] FIG. 1.21A shows a photograph of the tank of FIG. 1.5A containing assembled test units and coolant liquid in the tank cavity before boiling.

[0068] FIG. 1.21B shows a photograph of the coolant liquid of FIG. 1.21A boiling.

[0069] FIG. 1.21C shows a perspective view of an example assembled test unit.

[0070] FIG. 1.22A shows the measured temperature and differential tank pressure as a function of time when the two-phase immersion cooling system of FIG. 1.21 A operates at a static heat load of 600kW.

[0071] FIG. 1.22B shows the measured differential tank pressure as a function of time when the two-phase immersion cooling system of FIG. 1.21 A operates with a variable heat load.

[0072] FIG. 1.23 shows a shows a cross-sectional view of the tank of FIG. 1.4A corresponding to the plane A- A of FIG. 1 ,4B.

[0073] FIG. 1.24A shows a perspective view of a sensor bar in the tank of FIG. 1.23. The sensor bar includes a terminal box with a cover. The cover is transparent for the purposes of showing the components disposed within the terminal box.

[0074] FIG. 1.24B shows another perspective view of the sensor bar of FIG. 1.24A.

[0075] FIG. 1.24C shows a magnified perspective view of an electrical connector of the sensor bar of FIG. 1.24A.

[0076] FIG. 1.25 A shows a top, rear, right-side perspective view of the tank of FIG. 1.2A.

[0077] FIG. 1.25B shows a magnified right-side view of the tank of FIG. 1.25A where pressure sensors are located on the tank wall.

[0078] FIG. 1.26A shows example states for a programmable logic controller (PLC) for the two- phase immersion cooling system of FIG. 1.1 A.

[0079] FIG. 1.26B shows example transitions between the PLC states of FIG. 1.26A.

[0080] FIG. 1.26C shows a table of conditions that can trigger changes in the PLC states of FIG. 1.26A.

[0081] FIG. 1.27 shows a perspective view of the bellows in the two-phase immersion cooling system of FIG. 1.1 A.

[0082] FIG. 1.28 shows a cross-sectional perspective view of a pressure relief valve in the two- phase immersion cooling system of FIG. 1.1 A.

[0083] FIG. 1.29 shows the measured pressure across the pressure relief valve of FIG. 1.28 as a function of time during a pressure decay test.

[0084] FIG. 1.30 shows a cross-sectional view of the tank of FIG. 1.2A with a diffuser to reduce vapor movement at the top of the tank when the lid is in the open position.

[0085] FIG. 1.31 A shows a top perspective view of the lid on the tank of FIG. 1.2A.

[0086] FIG. 1.3 IB shows a bottom perspective view of the lid of FIG. 1.31 A.

[0087] FIG. 1.31C shows a cross-sectional view of the lid of FIG. 1.31 A corresponding to the plane A-A of FIG. 1.31 A. The lid includes a lid gasket, which is shown in an undeformed state.

[0088] FIG. 1.31D shows a cross-sectional view of the lid of FIG. 1.31A corresponding to the plane A-A of FIG. 1.31 A. The lid gasket is shown in a deformed state due to physical contact with a top plate of the tank.

[0089] FIG. 1.32 shows a cross-sectional view of the lid gasket of FIG. 1.31C.

[0090] FIG. 1.33 A shows a contour map of deformation across the lid gasket cross-section of FIG. 1.32 when the lid gasket is in the undeformed state of FIG. 1.31C. The contour plot is obtained using FEA.

[0091] FIG. 1.33B shows a contour map of deformation across the lid gasket cross-section of FIG. 1.32 when the lid gasket is in the deformed state of FIG. 1.3 ID. The contour plot is obtained using FEA.

[0092] FIG. 1.33C shows a contour map of deformation across a corner portion of the lid gasket of FIG. 1.32.

[0093] FIG. 1.33D shows an overlay of the lid gasket cross-section in its original undeformed state and in the deformed state at the plane A-A of FIG. 1.33C.|0094] FIG. 1.34 shows a table of various contaminants and options to reduce or, in some instances, mitigate each contaminant.

[0095] FIG. 1.35A shows a first part of a cable bill of materials for a representative assembled unit.[0096| FIG. 1.35B shows a second part of the cable bill of materials of FIG. 1.35A.

[0097] FIG. 1.36A shows an example tank with desiccant and a filtration system to remove contaminants from the coolant liquid disposed within the tank.

[0098] FIG. 1.36B shows a table of parameters related to the desiccant and the filtration system and their desired attributes.

[0099] FIG. 1.37A shows a diagram of an example filtration system.

[0100] FIG. 1.37B shows a top, front, left-side perspective view of the tank of FIG. 1.2A where cabinet doors are removed to show a filtration system (also referred to herein as an “external filter assembly”) disposed in a corresponding cabinet.|0101] FIG. 1.35C shows the components of the filtration system of FIG. 1.37B.

[0102] FIG. 1.38A shows a top perspective view of multiple desiccant assemblies disposed on a top plate of the tank of FIG. 1 ,2A. The tank is shown with the lid in the open position.

[0103] FIG. 1.38B shows a magnified view of a mounting bracket and an inner lid of the desiccant assembly of FIG. 1.38A.

[0104] FIG. 1.38C shows a top perspective view of the desiccant assemblies of FIG. 1.38A with respective inner lids removed.

[0105] FIG. 1.38D shows another top perspective view of the desiccant assemblies of FIG. 1.38C.

[0106] FIG. 1.38E shows a cross-sectional view of the tank and the desiccant assemblies of FIG.1.38A.

[0107] FIG. 1.39A shows a top perspective view of multiple desiccant assemblies of FIG. 1.36A.

[0108] FIG. 1.39B shows a top perspective view of one desiccant assembly of FIG. 1.39A with the inner lid removed.

[0109] FIG. 1.39C shows a side perspective view of a cage of the desiccant assembly of FIG. 1.39B.

[0110] FIG. 1.39D shows a perspective view of a container of the desiccant assembly of FIG. 1.39A.

[0111] FIG. 1.39E shows a perspective view of the container of FIG. 1.39D where a container cover is separate from a container base.

[0112] FIG. 1.40A shows a first set of components in the two-phase immersion cooling system of FIG. 1.1 A and example methods for removing and / or replacing the components.

[0113] FIG. 1.40B shows a second set of components in the two-phase immersion cooling system of FIG. 1.1 A and example methods for removing and / or replacing the components.

[0114] FIG. 1.41 A shows example validation criteria for various components of the two-phase immersion cooling system of FIG. 1.1 A.

[0115] FIG. 1.41B shows example validation criteria for various systems of the two-phase immersion cooling system of FIG. 1.1 A.

[0116] FIG. 1.42 shows the composition of an example coolant liquid for the two-phase immersion cooling system of FIG. 1.1 A.

[0117] FIG. 1.43 shows a cross-sectional perspective view of the tank of FIG. 1.2A with assembled units A and B inserted into the tank cavity via the guide rails.

[0118] FIG. 1.44 shows a magnified view of the assembled unit A of FIG. 1.43.

[0119] FIG. 1.45A shows a top view of the assembled unit B of FIG. 1.43.

[0120] FIG. 1.45B shows a magnified view of a mounting tab on the assembled unit B of FIG. 1.45A.[0121| FIG. 1.46 shows an example lift system to facilitate insertion and / or removal of assembled units to / from the tank of the two-phase immersion cooling system of FIG. 1.1A.

[0122] FIG. 1.47A shows an example end effector to mechanically engage an assembled unit to facilitate insertion and / or removal.

[0123] FIG. 1.47B shows an example gripper motor of the end effector of FIG. 1.47A.

[0124] FIG. 1.48 shows a perspective view of the end effector mechanically engaged to an assembled unit B.

[0125] FIG. 1.49 shows a top view of an example lift system to facilitate insertion and / or removal of assembled units to / from two two-phase immersion cooling systems.

[0126] Figure 2.1 shows a sleeve or gasket for protecting a server when it is installed in and removed from an immersion tank.

[0127] Figure 2.2A shows a front view of the sleeve of FIG. 2.1 attached to one side of a server chassis.

[0128] Figure 2.2B shows a top view of a server chassis with sleeves attached to both sides.

[0129] Figure 2.2C shows a detailed cross-section of a sleeve attached to one side of a server chassis.

[0130] Figure 2.3A shows a cross-sectional view of a server chassis with a sleeve attached to one edge installed in an immersion tank.

[0131] Figure 2.3B shows a close-up view of the server chassis and the sleeve between guide rails of the immersion tank.

[0132] Figure 2.3C shows a server chassis with sleeves on both edges installed in an alignment plate at the bottom of the immersion tank.[01331 Figure 2.3D is a close-up view of the server and sleeves installed in the alignment plate at the bottom of the immersion tank.

[0134] Figure 2.4 shows different views of an alternative sleeve for an immersion-cooled server chassis.E. DETAILED DESCRIPTION

[0135] Section 1: Two-Phase Immersion Cooling System Following below are more detailed descriptions of various concepts related to, and implementations of, a two-phase immersion cooling system for high power computing systems (e.g., computing systems having a power density greater than or equal to 250 kW per enclosure). It should be appreciated that various concepts introduced above and discussed in greater detail below may be implemented in multiple ways. Examples of specific implementations and applications are provided primarily for illustrative purposes so as to enable those skilled in the art to practice the implementations and alternatives apparent to those skilled in the art.

[0136] The figures and example implementations described below are not meant to limit the scope of the present implementations to a single embodiment. Other implementations are possible by way of interchange of some or all of the described or illustrated elements. Moreover, where certain elements of the disclosed example implementations may be partially or fully implemented using known components, in some instances only those portions of such known components that are necessary for an understanding of the present implementations are described, and detailed descriptions of other portions of such known components are omitted so as not to obscure the present implementations .

[0137] In the discussion below, various examples of a two-phase immersion cooling system are provided, wherein a given example or set of examples showcases a tank, alignment features (e.g., to guide the insertion and / or removal of a computing system), cable routing and passthroughs, a busbar, a power rack, a condenser, a sensor bar, a programmable logic controller (PLC), bellows, a pressure relief valve, a lid gasket, a filtration system (e.g., to purify the coolant liquid), a desiccant assembly, a lift system (e.g., to facilitate insertion and / or removal of an assembled unit to / from the tank), and a coolant liquid. It should be appreciated that one or more features discussed in connection with a given example of an immersion cooling system may be employed in other examples of immersion cooling systems, respectively, according to the present disclosure, such that the various features disclosed herein may be readily combined in a given immersion cooling system according to the present disclosure (provided that respective features are not mutually inconsistent).

[0138] Certain dimensions and features of the immersion cooling system are described herein using the terms “approximately,” “about,” “substantially,” and / or “similar.” As used herein, the terms “approximately,” “about,” “substantially,” and / or “similar” indicates that each of the described dimensions or features is not a strict boundary or parameter and does not exclude functionally similar variations therefrom. Unless context or the description indicates otherwise, the use of the terms “approximately,” “about,” “substantially,” and / or “similar” in connection with a numerical parameter indicates that the numerical parameter includes variations that, using mathematical and industrial principles accepted in the art (e.g., rounding, measurement or other systematic errors, manufacturing tolerances, etc.), would not vary the least significant digit.1. 1.1. An Example Immersion Cooling System

[0139] The two-phase immersion cooling systems disclosed herein (referred to hereafter as the “immersion cooling systems”) are configured to support high power computing systems, i.e., a computing system with a power density greater than 250 kW per enclosure (e.g., per tank) at peakpower. For example, the immersion cooling system may support a computing system with a power density of about 576 kW or about 600 kW. More generally, the immersion cooling systems disclosed herein may support computing systems having a peak power density ranging from about 250 kW to about 1000 kW per enclosure, including all values and sub-ranges in between. More preferably, the immersion cooling systems disclosed herein may support computing systems having a peak power density ranging from about 500 kW to about 1000 kW, including all values and sub-ranges in between. Herein, the power density is defined per enclosure (i.e., per tank). Additionally, the peak power corresponds to an upper limit in the power generated by the computing system contained within the tank. In other words, the computing system may generally have a power density that ranges from 0 kW to the peak power during operation. It should also be appreciated that the computing system may sustain the peak power over a prolonged period of time (e.g., more than 1 minute, more than 1 hour, more than 1 day).

[0140] Additionally, the immersion cooling systems disclosed herein may be configured to efficiently use electrical power and water. For example, the power usage effectiveness (PUE) of the immersion cooling system may range from about 1 to about 1.3, including all values and subranges in between. In some implementations, the power usage effectiveness (PUE) of the immersion cooling system may range from about 1 to about 1.15, including all values and sub-ranges in between. In another example, the water usage effectiveness (WUE) of the immersion cooling system may be less than or equal to about 0.5 L / kWh, including all values and sub-ranges in between. In some implementations, the water usage effectiveness (WUE) of the immersion cooling system may be less than or equal to about 0.2 L / kWh, including all values and sub-ranges in between.

[0141] In some implementations, the computing system installed in the immersion cooling system may include one or more assembled units. The assembled units may include, but are not limited to, servers, dedicated processing units, data-mining units, high bandwidth memory units, and other types of processing and / or memory units. Each assembled unit may include one or more printed circuit boards (PCBs) and / or one or more printed circuit board assemblies (PCBAs) with multiple surface-mounted electronic devices. The electronic devices include, but are not limited to, a computer processor unit (CPU), a graphics processor unit (GPU), a data processing unit (DPU), a field- programmable gate array (FPGA), an application-specific integrated circuit (ASIC), memory (e.g., rapid access memory), and, more generally, any electronic component with an integrated circuit. These surface-mounted electronic devices may be mounted to the PCB or PCBA using electrical interconnects, such as a ball grid array (BGA), a pin grid array (PGA), a land grid array (LGA), and / or the like.

[0142] FIGS. 1.1 A and 1.1B show an example two-phase immersion cooling system. As shown, the immersion cooling system includes a tank configured to contain coolant liquid and a computing system submerged in the coolant liquid. Specifically, the coolant liquid and the computing system may be disposed in a lower portion of a tank cavity defined by the tank. The tank further includes a lid, which may be opened to gain access to the tank cavity. For example, the lid may be opened to facilitate insertion and / or removal of one or more components of the computing system (e.g., one or more assembled units). Otherwise, during operation, the lid is closed, thus sealing the tank cavity.

[0143] The tank further includes a pair of condensers (see, for example, FIGS. 1.2H-1.2J) disposed, in part, in an upper portion of the tank cavity. The condensers are configured to extract heat from coolant vapor generated from the coolant liquid during operation, thus causing the coolant vapor to condense back to a liquid and fall into the pool of coolant liquid located in the lower portion of the tank cavity. Each condenser may extract heat from the coolant vapor using a fluid, such as water, circulated by an external fluidic system coupled to the tank. For example, each condenser includes one water box fluidically coupled to an inlet pipe to receive water at a relatively cold temperature from an external water source (e.g., a water supply system of a building) and another water box coupled to an outlet pipe to discharge water at a relatively hot temperature to the external water source. The pair of water boxes are fluidically coupled together via a plurality of fluid conduits (see, for example, FIGS. 1.2H-1.2J) disposed within the tank cavity.

[0144] The immersion cooling system further includes bellows coupled to a top plate of the tank via multiple venting conduits coupled to respective ports of the tank and the bellows. The bellows provides a way to regulate the tank pressure during operation by providing an additional space to store gas from the tank when the tank pressure increases and / or to supply gas to the tank when the tank pressure decreases.

[0145] Additionally, the immersion cooling system may include one or more pressure relief valves disposed on the tank and / or the bellows. Each pressure relief valve may be configured to open when the pressure difference across the pressure relief valve exceeds a pressure threshold associated with that valve, thus allowing fluid to pass between the tank cavity and the ambient environment. In some implementations, the pressure relief valve may be configured to open only when tank pressure is greater than the ambient pressure, thus reducing or, in some instances, mitigating overpressure-related failures. In some implementations, the pressure relief valve may be configured to open only when ambient pressure is greater than the tank pressure, thus reducing or, in some instances, mitigating negative pressure-related failures.

[0146] The immersion cooling system may further include a pair of power racks separate from the tank to supply electrical power to the computing system disposed within the tank and / or other components requiring electrical power in the immersion cooling system. Each power rack may receive electrical power from an external power source. For example, the power rack may receive an alternating current (AC) feed from an overhead low voltage (LV) panel in a building. Each power rack may further supply electrical power to the tank via a busbar (e.g., the busbar A in FIG. 1.1 A). The busbar may, in turn, be disposed towards the bottom side of the tank cavity (see, for example, FIGS. 1.2H and 1.21) and the various components of the computing system may be electrically coupled to the busbar to receive electrical power upon installation into the tank.

[0147] By using the power racks and the busbars to provide electrical power, the components of the computing system (e.g., an assembled unit) may be installed without any onboard power supply. This, in turn, reduces the volume occupied by the computing system and / or allows placement of additional electronic components within the tank (e.g., an additional processor, additional memory). Additionally, the pair of power racks and the pair of busbars are disposed on opposing sides of the tank. This configuration reduces the net impedance of the busbar (e.g., by reducing the distance from the power rack to the end of the busbar located furthest from the power rack) and provides a way to more easily service and / or replace the busbar.

[0148] The immersion cooling system further includes a pair of grounding assemblies disposed on outer tank walls of the tank. Each grounding assembly provides a ground connection for the tank. As shown, the grounding assembly may include one or more ground cables electrically coupled to a bracket mounted to a tank wall of the tank.

[0149] FIGS. 1.2A-1.2J show additional views of the tank in the immersion cooling system of FIG. 1.1 A. The tank includes a tank frame supporting multiple tank walls that define the tank cavity. The tank frame may further include a top plate defining a top opening through which the computing system may be inserted and / or removed from the tank cavity (see Section 1.2 and Section 1.10 for further details). For example, FIG. 1.3 shows a cross-sectional view of the tank where multiple assembled units of the computing system are inserted into the tank cavity and electrically coupled to a busbar A. The tank further includes a lid that may be moved between an open position to provide access to the tank cavity through the top opening and a closed position to seal the tank cavity from the ambient environment. The lid may be securely closed via a locking mechanism. For example, the locking mechanism on the lid shown in FIGS. 1.2A-1.2J may include a spring- loaded latch mechanism.

[0150] As described above, the tank cavity may be divided into a lower portion where the computing system and the coolant liquid are located and an upper portion where a pair of condensers arelocated. During operation, the coolant liquid may be located just below the condensers (e.g., the lowest portion of the conduits within the tank cavity). Thus, the vapor space (e.g., the space above the coolant liquid that predominantly contains coolant vapor) and the head space above the vapor space (e.g., the space that contains coolant vapor and air) may be predominantly located within the upper portion of the tank cavity.

[0151] The tank may be relatively compact in size to facilitate installation into a building space while maintaining sufficient volume to contain coolant liquid and the computing system. For example, the tank may have a length of about 2 m. More generally, the length may range from about 2 m to about 3m, including all values and sub-ranges in between. The tank may have a width of about 1 m. More generally, the width may range from about 0.5 m to about 1.5 m, including all values and sub-ranges in between. The tank may have a height of about 1.5 m. More generally, the width may range from about 1 m to about 2 m, including all values and sub-ranges in between. The lower portion of the tank cavity may have a width of about 0.8 m. More generally, the width of the lower portion may range from about 0.5 m to about 1.5 m, including all values and subranges in between. The tank cavity may be configured to contain coolant liquid at a height defined from the bottom of the tank cavity to the surface of the coolant liquid of at least 0.7 m. More generally, the width of the height of the coolant liquid may range from about 0.7 m to about 1 m, including all values and sub-ranges in between. It should be appreciated that the foregoing dimensions are non- limiting and that smaller and larger tanks with different- sized tank cavities are also contemplated in the present disclosure.|0152] Each condenser may include a pair of water boxes mounted to an exterior portion of the tank - one water box to receive cold water for the immersion cooling system and another water box to discharge hot water from the immersion cooling system. The water boxes may be fluidically coupled together via a plurality of fluid conduits. In some implementations, the water boxes and the fluid conduits may be coupled such that the water flows through the tank cavity multiple times. In other words, the condenser may be configured to operate as a multipass heat exchanger (see Section 1.4 for further details). It should be appreciated that water is a non-limiting coolant for the condenser. More generally, the condenser may use various coolants to extract heat from the coolant vapor including, but not limited to, water (e.g., deionized water), glycol (e.g., ethylene glycol, propylene glycol), and a water / glycol solution.

[0153] The tank further includes multiple ports disposed on the top plate to fluidically couple the tank to the bellows via the venting conduits. FIG. 1.2B further shows the top plate may include a pressure relief valve. Further details on the bellows and the pressure relief valve may be found in Section 1.6.

[0154] The tank may further provide multiple busbar feedthroughs to facilitate insertion of a busbar from a power rack into the tank cavity. As shown in FIGS. 1.2F and 1.2G, the right and left sides of the tank may each include a pair of busbar feedthroughs - an upper busbar feedthrough collocated with the upper portion of the tank and a lower busbar feedthrough collocated with the lower portion of the tank. The upper and lower busbar feedthroughs provide a way to accommodate different busbars and / or different power racks. In FIGS. 1.2A-1.2J, a pair of busbars A are inserted through respective upper busbar feedthroughs on each side of the tank. For completeness, FIGS. 1.4A and 1.4B show another example where the tank includes a pair of busbars B inserted through respective lower busbar feedthroughs on each side of the tank. The busbars A and B include respective base bars disposed on the bottom surface of the tank cavity with multiple electrical sockets that electrically couple to corresponding electrical pins on assembled units. In some implementations, the upper busbar feedthrough may reduce the likelihood of a coolant liquid leak since the upper busbar feedthrough is collocated with the portion of the tank cavity containing coolant vapor and air. Further details of busbars are provided in Section 1.3.

[0155] The tank may also include one or more cabinets to contain various auxiliary systems of the immersion cooling system. The auxiliary systems include, but are not limited to, control electronics and a filtration system to purify the coolant liquid (e.g., by removing contaminants) during operation. Further details on the filtration system and other fluid conditioning components may be found in Section 1.7.

[0156] The tank may further support various sensors to monitor the operating conditions of the immersion cooling system. The operating conditions include, but is not limited to, the tank pressure, the temperature of the coolant liquid, the tempreature of the coolant vapor / air, the height of the coolant liquid within the tank, the relative humidity of the coolant vapor / air, and the like. The data acquired by the various sensors in the immersion cooling system may be transmitted to a programmable logic controller (PLC) (not shown) and a controller (not shown). The controller may be a “tank manager” controller configured to control operations of the tank or components inside the tank such as condenser coil flow rate, immersion cooling filter flow rate, server racks or semiconductor dies including throttling, rate limiting, or depowering semiconductor dies within an immersion cooling system, ventilation systems, cooling fans, tank power, and the like. The PLC may be limited to simple functions as compared to controller. These functions may include monitoring the one or more sensors, disconnecting power to tank, changing to one of a predetermined number of states, and sending notification messages (also referred to as tickets) to an entity such as a human operator or controller. The benefit of this simple functionality is increased robustness, reliability, and resistance to environmental disruptions such as power surges, abnormally high orlow temperatures, vibrations (e.g., due to seismic activity) as compared to more general-purpose computers. PLC may also provide improved resistance to malware, hacking, and similar digital attack vectors designed to disrupt functioning of systems to which PLC is communicatively coupled when compared to controller or similarly complex control devices. Further details on sensors, the PLC, and the controller may be found in Section 1.5.

[0157] FIGS. 1.5A-1.5E show several photographs of an example demonstration of a two-phase immersion cooling system that incorporates the various inventive concepts disclosed herein.2. 1.2. Installation Features of the Two-Phase Immersion Cooling System

[0158] The tank of the immersion cooling system may include various alignment features to facilitate the insertion and removal of components of the computing system (e.g., the assembled units) to / from the tank. For example, FIGS. 1.6A-1.6C show the tank cavity may include multiple guide rails to guide an assembled unit through the upper portion of the tank cavity. In this manner, the guide rails may reduce or, in some instances, prevent damage to components of the immersion cooling system located in the upper portion of the tank cavity (e.g., the condenser, a sensor bar) by reducing or, in some instances, preventing physical contact between the assembled unit and aforementioned components of the immersion cooling system. A further example would include a multi-purpose sleeve or gasket for protecting the server when it is installed and removed from an immersion tank for the immersion cooling system.

[0159] FIGS. 1.7A-1.7E show the tank may further include a tray disposed on the bottom of the tank cavity that forms a channel containing the base bars of the busbar A (or busbar B). The tray may be coupled to the tank via one or more fasteners. The assembled unit may couple to a busbar (e.g., the busbar A or B) by inserting a set of electrical pins on the assembled unit into corresponding electrical sockets on the busbar. To guide the insertion of the electrical pins of the assembled unit into the electrical sockets of the busbar, the assembled unit and the tray may include guide pins and guide sockets, respectively. For example, FIG. 1.8 shows a bottom portion of an example assembled unit with multiple electrical pins and guide pins. In FIGS. 1.7A-1.7E, the electrical pins and the guide pins are shown inserted into respective electrical sockets of the busbar A and respective guide sockets of the tray.

[0160] In some implementations, the combination of the guide rails and the guide pins / sockets may provide both coarse and fine alignment of the assembled unit to the tank. For example, the guide rails may be dimensioned and / or located within the tank cavity to provide coarse alignment. The guide pins / sockets provide fine alignment and support for the assembled unit. In this manner,the assembled units may be mechanically constrained from the top of the tank cavity to the bottom of the tank cavity where the coolant liquid is located. In some implementations, the guide rails may be arranged such that the bottom of the assembled unit "floats" so that the busbar guide pins can provide tight alignment. This approach of providing both coarse and fine alignment may reduce or, in some instances, mitigate alignment errors between the assembled unit and the tank caused by manufacturing and assembly tolerances of the tank and / or variations in the dimensions of the tank during operation.

[0161] The assembly tolerances between the busbar and the tank may be reduced using an alignment fixture. For example, FIG. 1.9 shows an example alignment fixture, which may be inserted into the tank cavity to mechanically constrain the placement of the busbar with respect to the tank. In some implementations, the alignment fixture may also function as a no-go gauge for busbars and tank openings.

[0162] To further reduce or, in some instances, mitigate alignment errors - particularly between the assembled unit and the guide rails, a compliant gasket (also referred to as a sleeve) may be disposed on the sides of the assembled unit to fill any gaps between the assembled unit and each guide rail. The sleeve allows for “less-than-perfect” mating of the server chassis to guide rails and / or alignment features within the immersion tanks. The sleeve prevents metal-on-metal contact between the server chassis and the guide rails within the immersion tank, reducing the probability of creating metal debris when inserting or removing the server chassis. And the sleeve occupies space inside the immersion tank that would otherwise be occupied by the immersion fluid, reducing the quantity of immersion fluid needed to fill the immersion tank. Further details and examples of a sleeve to facilitate installation and / or removal of an assembled unit to / from the tank may be found in U.S. Provisional Application No. 63 / 516,996, filed on August 1, 2023 and entitled, “Multi-Purpose Sleeve for Immersion-Cooled Servers,” which is incorporated by reference herein in its entirety.

[0163] Additionally, the tank may be designed such that the lower portion of the tank exhibits relatively low deformation during operation. For example, FIGS. 1.12A and 1.12B show FEA simulation results of the deformation for a portion of the tank when the tank cavity is at a gauge pressure of -100 mbar. As shown, the deformation within the lower portion of the tank cavity is less than 4 mm. Additional simulation results show deformation is less than 3mm at a gauge pressure of -69 mbar (-1 psi).

[0164] Moreover, the foregoing components of the tank and the busbar may be manufactured with specified tolerances to ensure the guide rails, the guide pins / sockets, and the electrical pins / sockets are approximately aligned during assembly. For example, the guide pins and the electrical pins ofan assembled unit may have tolerances of + / - 0.5 mm and + / -0.9 mm, respectively. In another example, the tray may have tolerances of + / - 0.5 mm. The width of the top opening of the tank may have tolerances of + / - 4 mm. The studs on the bottom of the tank cavity where the tray is mounted may have tolerances of + / - 1 mm. The guide rails may have tolerances of + / - 1 mm.[0165| In some implementations, the guide pins may have a diameter of about 12 mm. The guide sockets may be tapered to provide up to 10 mm of insertion misalignment. In some implementations, the electrical sockets of the busbar may float by about 1 mm so that any mechanical loads from the assembled unit are supported by the guide pins and the tray. The guide pins may be offset spatially on the assembled unit to prevent backward installation of the assembled unit to the tank.

[0166] The tray may accommodate different busbars (e.g., the busbars A and B). For example, the tray may provide a sufficiently large channel to accommodate the base layers of different busbars.

[0167] In some implementations, the guide rails, the guide sockets, and the electrical sockets may be distributed along the length of the tank at equal increments. The increments may be based on standardized dimensions commonly used in server racks, such as a rack unit. For example, the increments may be equal to 1U (1.75 inches), 2U (3.5 inches), 3U (5.25 inches), and so on. In some implementations, the placement and distribution of the guide rails may be varied to accommodate different sized assembled units. For example, the guide rails may include at least one pair of guide rails to accommodate lU-sized assembled units and at least one pair of guide rails to accommodate 2U-siezd assembled units. Moreover, the guide rails may be readily replaced and / or moved within the tank cavity to accommodate different computing systems as desired.

[0168] As described above, the immersion cooling system may support multiple assembled units. The number of assembled units may vary depending on the size of each assembled unit. For example, the number of 2U assembled units in the tank may range from 32 to 64, including all values and sub-ranges in between.

[0169] The computing system installed in the tank may require multiple cable connections for operation. The cable connections may include, for example, fiber optic cables and / or copper wire cables. The cables may be arranged using a cable routing bracket located between the top plate and the condenser, as shown in FIG. 1.10. In particular, the cable routing bracket may provide a channel to route cables to multiple passthroughs located on the tank wall that each have a passthrough module mounted thereto (e.g., a module to facilitate connection to cables located within the tank cavity and outside the tank).

[0170] In one illustrative example, each assembled unit may have six network cables, e.g., four 800 Gb, 16F octal small form factor (OSFP) cables that connect to Cedar modules, and two 400Gb,8F quad small form factor dual density (QSFP-DD) cables that connect to BF3 modules. The computing system may further include 64 assembled units, resulting in 384 cables in total. The 384 cables may be routed to 12 passthrough modules located on both left and right sides of the tank.

[0171] In some implementations, the tank may include multiple cable routing brackets and the number of cables may be divided accordingly. Referring to the above example of a computing system with 384 cables, the tank may include four cable routing brackets and the cables may be divided into quarters accordingly with 96 cables routed in each bracket. The passthrough modules may support 432 cable connections resulting in 12% spare capacity. The passthrough modules may further be adaptable to externally structured cabling.

[0172] FIG. 1.11 shows an example passthrough module for multiple fiber optic connections. In one example, the passthrough module may support 36 16-fiber very small form factor (VSFF) multi-fiber push on (MPO) connectors allowing for point-to-point connections between the assembled units and the passthrough module. In another example, the passthrough module may support three 192-fiber expanded beam optical (EBO) connectors. In both examples, the passthrough module supports 576 fiber structured cables. The passthrough module may be used for both front-end and back-end network connections. The passthrough module may include fiber ribbons that are epoxy sealed internally. The passthrough module may be sealed to the outside of the tank with a gasket. In some implementations, the passthrough module may have the following dimensions: HlOOmm x W78mm x D151mm. However, it should be appreciated that the dimensions are nonlimiting. In particular, the depth of the passthrough module may be reduced using shorter fiber optic pigtails (e.g., D may be about 25 mm).3. 1.3. Power Supply and Distribution

[0173] As described above, the immersion cooling system may include power racks to supply electrical power to the computing system. This is facilitated through the use of busbars to transmit electrical power from the power rack to the tank. For example, FIG. 1.13 shows the tank has an upper busbar feedthrough for the busbar A and a lower busbar feedthrough for the busbar B. For illustration, both busbars A and B are shown mounted to their respective busbar feedthroughs. However, it should be noted that, in practice, only one busbar feedthrough is used. In some implementations, the upper busbar feedthrough may be preferable to use to reduce or, in some instances, mitigate loss of coolant liquid since the upper busbar feedthrough is collocated with the upper portion of the tank cavity where the head space is located.

[0174] The busbars A and B may be mounted to the tank using the same tray described above. Further, the busbars A and B may provide electrical sockets positioned at the same location in the tank cavity, thus providing support for the same assembled units.

[0175] This arrangement is distinguished over conventional computing systems, which typically include onboard power supplies located in close proximity to the electronic components of the computing system. The removal of the power supplies from the computing system thus allows space savings with the tank. For example, the tank cavity may have a smaller volume and, hence, use less coolant liquid for the same computing system. In another example, the space saved by the removal of the power supplies may be used to contain other electronic components to further increase the computational power of the computational system (e.g., an additional processor, additional memory).

[0176] However, offloading the power supplies to the power racks creates other challenges. In particular, transmitting dynamically variable currents (e.g., up to 6000 A) at low voltages across relatively long distances (e.g., greater than or equal to about 4 m between the endpoint of the busbar and the power rack) is highly non-trivial. For example, the bus bar typically supports multiple resonances, which can cause the supplied voltage to fluctuate in an undesirable manner in response to changes in the load of the computing system (e.g., the current supplied may vary up to 6000 A in a few microseconds). Conventional methods to dampen these resonances by adding resistance is impractical due to higher losses and difficulties in controlling the fast dynamics of the busbar.

[0177] The foregoing issue of busbar resonances is addressed herein, in part, by designing the bus bars to have appreciably low inductances. For example, FIGS. 1.14A and 1.14B show the busbar B includes a first base bar for power and a second base bar for return. The first and second base bars extend along the length of bus bar and bifurcates at the power rack, thus maintaining impedance along the length of the busbar. This configuration can achieve an end-to-end inductance of about 1 pH.

[0178] In another example, FIGS. 1.15A-1.15C show the busbar A includes a multilayer base bar with alternating layers providing power and return to reduce the magnetic B-field. For example, the busbar A may include four layers that each have a thickness of 6.35mm. A 0.5mm thick insulator may further be disposed between respective layers of the multilayer bus bar. Similar to the busbar B, the power and return layers of the busbar A may bifurcate at the power rack (see FIGS. 1.16A and 1.16E), thus maintaining a relatively constant impedance along the length of the busbar A. This configuration can achieve an end-to-end inductance of about 50 nH.

[0179] FIGS. 1.16A-1.16E show the busbar A mounted to the tank via a feedthrough assembly. FIGS. 1.17A-1.17C show additional views of the feedthrough assembly separate from the tank. As mentioned above, the feedthrough assembly being located on the upper busbar feedthrough may appreciably reduce coolant liquid loss. Additionally, the feedthrough assembly may provide a mechanically compliant seal, thus decoupling the busbar A from deformation in the tank wall during operation (e.g., tank wall flex), which may arise due to fluctuations in the tank pressure. This is accomplished, in part, by the feedthrough assembly including a flexible section that joins a first flange coupled to the tank wall and a second flange coupled to the busbar (e.g., the external multilayer power bar). The second flange may be coupled to the busbar via an epoxy seal. FIG. 17C further shows the multilayer power bar may include one or more milled channels to facilitate infiltration of epoxy during assembly. The flexible section may thus reduce or, in some instances, mitigate transfer of any mechanical load from the tank wall to the epoxy seal.

[0180] In some implementations, the feedthrough assembly may provide a near-unity pressure ratio due to the compliance of the flexible section. Additionally, any stress at the epoxy-busbar joint may be limited to thermal effects (e.g., thermal expansion, contraction).

[0181] Additionally, the busbar A is adjustable at its joints and the rack to facilitate installation to the tank and the power rack. Further, the flexible section is compliant to accommodate offset in the alignment of the power rack and the tank. The passthrough block may be attached to the tank and the power rack via lap joints. Once all fasteners for the feedthrough seal are positioned, a final torque is applied to couple the feedthrough assembly and, hence, the busbar to the tank. Thereafter, the flexible section may remain fixed in place to the tank, but still remain compliant.

[0182] The formation and assembly of the epoxy seal may be an off-line manufacturing process, which allows use of high-control processes (e.g., vacuum potting). This allows for greater repeatability for manufacture and assembly. Additionally, the epoxy seal may be field-replaceable even with the tank loaded and filled with coolant liquid and a computing system.

[0183] The busbar A is assembled via bolts and a full-lap joint on each layer. The respective layers of the multilayer bus bar may not be adhered to one another. The flexible section of the feedthrough assembly may be compressed temporarily during assembly. The coolant liquid may condense in the flexible section but when the computing system is under load, the busbar will be at a temperature greater than or equal to the boiling point of the coolant liquid (e.g., 49°C).

[0184] The busbar A may include a pin and socket layout defined according to the Sentrality System. The electrical sockets of the busbar A may electrically couple individual layers to the assembled units. For example, the assembled unit may include four electrical pins with different lengthsto couple the pins to the four layers of the multilayer bus bar. The busbar A may further have a blind-mate design (e.g., using the pins and sockets). The electrical sockets may provide 1 mm of float, as described above. Insulating bushings may be included to reduce the likelihood of a pinbar short.

[0185] Further details and examples of a busbar may be found in U.S. Provisional Application No. 63 / 514,259, filed on July 18, 2023 and entitled, “Busbar for Immersion-Cooling System,” and U.S. Provisional Application No. 63 / 515,423, filed on July 25, 2023 and entitled, “Busbar for Immersion-Cooling System.” Each of the aforementioned references is incorporated by reference herein in its entirety.

[0186] FIG. 1.18 shows a diagram of the various electrical connections between the power racks and the components of the tank in the two-phase immersion cooling system of FIG. 1.1 A and / or the computing system. FIG. 1.19 shows an example circuit diagram for the two-phase immersion cooling system of FIG. 1.1 A.

[0187] The power racks in the immersion cooling system may be communicatively coupled to the PLC via a cable (e.g., an umbilical). This provides a way for the PLC to shut down power via an emergency stop (E-STOP). Additionally, the E-STOP may reach-back to low voltage (LV) distribution in the facility and cut corresponding breakers. As described in Section 1.5, the PLC may provide state monitoring for safety and to identify faults in the immersion cooling system while the controller provides full telemetry. The control plane may further interface with the controller.

[0188] In some implementations, the emergency stop command is hardware-only. This eliminates any possibility of a software update-derived bug affecting the emergency stop function. After an emergency stop. After an E-STOP occurs, the immersion cooling system may require the control plane to restart. The open umbilical may also require a software trigger to restart. The PLC and power rack connection may provide high reliability. For example, electrical power may still be provided if the PLC fails. The PLC may be fed 2x N+l using an uninterruptable power supply (UPS)-backed AC power supply.4. 1.4. A Multipass Condenser

[0189] The condensers in the immersion cooling system of FIG. 1.1 A may generally provide sufficient cooling to support high power computing systems with a power density greater than 250kW. In some implementations, the condensers may provide at least 576kW of cooling using water at a flow rate of 300 gallons per minute (GPM) and an inlet fluid temperature of 37°C. The condensers may further provide this cooling capacity when the lid of the tank is in both the closed positionand the open position (e.g., when the immersion cooling system is being serviced). In some implementations, the pressure drop across the condenser may be less than 200 mbar when the flow rate is 300 GPM.

[0190] It should be appreciated that the foregoing parameters of the condenser are non-limiting examples and may vary depending on the power density of the computing system. Generally, the cooling power may be equal to or greater than the power density of the computing system. For example, the condensers may provide a cooling power ranging from about 250 kW to about 1000 kW, including all values and sub-ranges in between. In another example, the condensers may support a flow rate of water ranging from about 1 GPM to about 500 GPM, including all values and sub-ranges in between. In another example, the temperature of the inlet fluid temperature may range from about 20°C to about 40°C, including all values and sub-ranges in between.|0191] FIGS. 1.20A-1.20C show several views of the condensers. As shown, each condenser may include a pair of water boxes mounted to an exterior portion of the tank - one water box to receive cold water for the immersion cooling system and another water box to discharge hot water from the immersion cooling system. The water boxes may further be fluidically coupled together via a plurality of condenser conduits (also referred to as “condenser tubes”).

[0192] In some implementations, the condensers may be configured to operate as a multipass heat exchanger. For example, FIGS. 1.20B and 1.20C show the water boxes include partitions that form multiple cavities. Each cavity may be coupled to a different set of condenser conduits. Thus, the water may flow between the water boxes multiple times via different condenser conduits before being discharged from the tank. For the condenser shown in FIGS. 1.20B and 1.20C, the water may flow three times between the water boxes before being discharged through the outlet (i.e., the condenser is a 3-pass condenser).

[0193] When the condensers operate with water flowing at a flow rate on the order of hundreds of gallons per minute, the water pressure may be appreciably high, thus increasing the risk of a water leak in the condenser. A large water leak may be readily detected by monitoring the water pressure through the condenser conduits. For smaller leaks, the condenser may include a leak detection system with pressure sensors disposed within one or both of the water boxes of each condenser. Further details and examples of a leak detection system for the condenser and, in particular, the water boxes, may be found in U.S. Provisional Application No. 63 / 519,095, filed on August 11, 2023 and entitled “EXTERNAL LEAK DETECTION SYSTEM,” which is incorporated by reference herein in its entirety.

[0194] To demonstrate the cooling capacity of the two-phase immersion cooling system disclosed herein, a cooling test was performed using the example two-phase immersion cooling system of FIG. 1.5A. Both static and dynamic heating conditions were evaluated. In particular, FIGS. 1.21A and 1.21B show photographs of the tank cavity in the two-phase immersion cooling system of FIG. 1.5A with assembled test units and coolant liquid before boiling and during boiling. This demonstration was performed using assembled test units with a 2U chassis that includes four sets of fin heaters, as shown in FIG. 1.21C. Each assembled test unit provided a 19.2 kW heat load. The immersion cooling system contained 32 assembled test units resulting in a total heat load of 614 kW. Additionally, the coolant liquid used is Unistar Unisolv-IC.

[0195] For the static load tests, the test was performed at a static heat load of 600kW, with the condenser fluid (water) at a temperature of 39°C and a flow rate of 300GPM. The coolant vapor pressure within the tank cavity was able to stabilize. The tank pressure was maintained close to ambient pressure so that the condenser performance value was evaluated under a worst-case scenario (i.e., the open-lid scenario during servicing). FIG. 1.22A shows the measured temperature of the water and the differential tank pressure as a function of time under these conditions. The static cooling test demonstrated the condensers are capable of providing 600 kW of cooling for a sustained period of time.

[0196] For the dynamic load tests, a power-cycle was performed. Specifically, the transitioned from “idle state,” to a 50kW heat load, and then to a ~ 530kW heat load (e.g., the highest orange line in FIG. 1.22B). FIG. 1.22B shows the measured differential tank pressure as a function of time when the two-phase immersion cooling system of FIG. 1.21 A operates with a variable heat load. The results in FIG. 1.22B show that when power was brought to 530kW from 50kW, the vapor pressure was able to stabilize to around 40mBar after an initial spike regardless of whether the prior idle time is 30min, or 3min.5. 1.5. Sensors & Programmable Logic Controller (PLC)

[0197] The immersion cooling system may include various sensors to monitor the operating conditions. The immersion cooling system may further include a (PLC) (not shown) and a controller (not shown). As described above, the controller may be a “tank manager” controller configured to control operations of the tank or components inside the tank such as condenser coil flow rate, immersion cooling filter flow rate, server racks or semiconductor dies including throttling, rate limiting, or depowering semiconductor dies within an immersion cooling system, ventilation systems, cooling fans, tank power, and the like. The PLC may be limited to simple functions as compared to controller. These functions may include monitoring the one or more sensors, monitoringthe health and sate of the power rack and / or the tank, disconnecting power to tank, changing to one of a predetermined number of states, and sending notification messages (also referred to as tickets) to an entity such as a human operator or controller. The benefit of this simple functionality is increased robustness, reliability, and resistance to environmental disruptions such as power surges, abnormally high or low temperatures, vibrations (e.g., due to seismic activity) as compared to more general-purpose computers. PLC may also provide improved resistance to malware, hacking, and similar digital attack vectors designed to disrupt functioning of systems to which PLC is communicatively coupled when compared to controller or similarly complex control devices.

[0198] The sensors integrated into the immersion cooling system may generally be serviceable and redundant. The sensors may include, but are not limited to, temperature sensors (e.g., to monitor fluid and head space temperatures, to provide data to determine the water dew point in the head space, to monitor the inlet fluid temperature of the condenser), humidity sensors (e.g., to measure relative concentration of water in the head space, to provide data to calculate the water dew point), pressure transducers (e.g., to measure / detect over / under pressurization events, to detect condenser failures (hard / soft), to monitor the inlet pressure of the condenser), fluid level sensors (low-mid-high) (e.g., to monitor critical low, top-off and critical high fluid levels, to issue maintenance tickets or assert critical power off), and a lid proximity sensor (e.g., to detect whether the lid is in the open position or closed position). In some implementations, the immersion cooling system - in particular, the PLC and controller - may also receive data from sensors deployed in the facility near the immersion cooling system, such as sensors to detect a coolant vapor leak.|0199] In some implementations, the sensors deployed within the tank cavity may be mounted to a sensor bar. As shown in FIG. 1.23, the sensor bar may be mounted to an interior side of the tank wall. FIGS. 1.24A-1.24C show additional views of the sensor bar separate from the tank. As shown, the sensor bar may include a rail to support various sensors and / or electrical connections to the sensors. The rail may further extend from the upper portion of the tank cavity to the lower portion of the tank cavity. Thus, the sensor bar may support sensors that monitor conditions of the coolant liquid and the head space. For example, FIGS. 1.24A and 1.24B show the rail may support temperature sensors and humidity sensors for the head space, coolant liquid level sensors to monitor the level of the coolant liquid, and temperature sensors to monitor the temperature of the coolant liquid. The sensor bar may further include a terminal box containing various terminal blocks for various electrical connections to the sensors.

[0200] The sensor bar may further include a connector assembly mounted to the rail and coupled to a passthrough on the tank. As shown in FIG. 1.24C, the connector assembly may include an electrical connector that provides electrical power and communication to the various sensors 1supported by the sensor bar. In some implementations, the electrical connector may be a DB25 connector. The electrical connector may be directly connected to the PLC during operation.

[0201] The tank may further include sensors for the condenser. For example, tempreature and pressure sensors may be mounted on water boxes of each condenser to monitor the temperature and pressure of the water at the inlet and outlet. Additionally, several sensors may be disposed on the exterior of the tank. For example, FIGS. 1.25A and 1.25B show several pressure transducers may be mounted onto the tank wall to monitor the tank pressure.

[0202] It should be appreciated that the sensors above are non-limiting examples. In some implementations, additional sensors may be included, for example, to measure the liquid level, froth level, boiling point of the coolant liquid, liquid impedance, froth impedance, vapor impedance, and / or dielectric constants of the coolant liquid. Further details and examples of these fluid sensors that may be found in U.S. Provisional Application No. 63 / 512,575, filed on July 7, 2023 and entitled, “Fluid Sensor for Immersion-Cooling Systems,” and U.S. Provisional Application No. 63 / 516,748, filed on July 31, 2023 and entitled, “DI-ELECTRIC MONITORING OF IMMERSION FLUID DURING COOLING OPERATION.” Each of the aforementioned references is incorporated by reference herein in its entirety.

[0203] The PLC may be programmed with a number of states which provide a context- specific framework for interpreting sensor readings and characterizing operational conditions of the immersion cooling system. These states may be represented by a state machine, illustrated in FIG. 1.26A. The state machine may include a collection of PLC states that allow the PLC to provide context-specific information to the controller based on the combination of sensor readings and PLC state. In some implementations, the PLC may be in one of eight PLC states: energize state, normal state, service state, lid open state, throttle state, soft off state, critical off state, and critical power state.

[0204] FIG. 1.26B shows a state transition map summarizing the transitions between the PLC states. The state transition map groups states into two categories of potentially “healable” and not “healable.” “Healable,” as used herein, indicates a PLC state that may transition back to normal state by itself without being forced back into normal state by controller or a human operator. For example, if the lid is opened, PLC will transition into the lid open state. Once the lid is closed again, the PLC will transition from the lid open state back into the normal state without any additional commands from the controller or the human operator. However, if the PLC transitions into a non-healable state (any of states in the dashed box), it will not be able to transition back into the normal state until the controller or the human operator issues an instruction for PLC to do so. FIG.1.26C provides a table summarizing some of the transitions from a PLC “normal” state and the conditions that may trigger these transitions.

[0205] Further details of a PLC for the two-phase immersion cooling systems disclosed herein may be found in U.S. Provisional Application No. 63 / 519,086, filed on July 7, 2023 and entitled, “LOCAL PLC CONTROLS WITH STATE MACHINE FOR OPERATIONAL CLOUD,” which is incorporated by reference herein in its entirety.6. 1.6. Coolant Vapor Management

[0206] The immersion cooling system of FIG. 1.1 A may further include multiple components to regulate and manage the generation of coolant vapor and the resulting changes to the tank pressure during operation.

[0207] For example, the immersion cooling system may include bellows (also referred to as a “bellows assembly”) coupled to the tank, as shown in FIG. 1.27. The bellows may be used to regulate the tank pressure by providing additional volume to receive gas (e.g., a mixture of coolant vapor and air) from the tank when tank pressure increases or to supply the gas to the tank when the tank pressure decreases. In some implementations, the bellows may regulate the tank pressure when the tank operates at or near atmospheric pressure (e.g., the ambient pressure of the environment surrounding the immersion cooling system). Additionally, the bellows may stabilize the tank pressure in the tank during startup, shutdown, and / or when the heat load is varied during operation. By providing an additional volume to contain coolant vapor, the bellows may also reduce loss of coolant liquid during operation as the tank pressure varies. In other words, without the bellows, the coolant vapor generated from the coolant liquid may be vented directly to the atmosphere when the tank pressure is too high.

[0208] The bellows may include a container with at least one flexible, polymeric wall that can deform outward from a center of the bellows and deform inward to change an amount of volume enclosed by the container. In some implementations, the inflatable container may be covered by a stamped aluminum outer housing. Various hardware elements (such as mounting hardware, pressure-relief vent(s), and / or port(s)) may be mounted on the container.

[0209] The bellows may have the following dimensions: 2.7m x 1.1m x 0.3m. It should be appreciated, however, the foregoing dimensions are non-limiting and the bellows may be shaped and / or dimensioned differently to accommodate different- sized tanks and / or to provide more or less volume to contain the gas.

[0210] The bellows may further include one or more pressure relief valves (see FIG. 1.28) disposed on a top portion of the outer housing. The pressure relief valve(s) may be configured to open when the pressure within the container is greater than or equal to 1 psi to alleviate the overpressure conditions within the bellows. It should be appreciated this pressure threshold is a non-limiting example and that it may be higher or lower depending on the immersion cooling system’ s tolerance to overpressure conditions. Generally, the coolant vapor has a higher density than air resulting in higher concentrations of coolant vapor towards the bottom of the bellows. By placing the pressure relief valves on the top of the outer housing, the loss of coolant vapor may be reduced in the event the pressure relief valves open.

[0211] In some implementations, the bellows may include a drain valve to empty the contents of the bellows, e.g., during servicing.

[0212] Further details and examples of bellows may be found in International PCT Application No. PCT / US2023 / 067056, filed on May 16, 2023 and entitled, “BELLOWS FOR IMMERSION COOLING,” which is incorporated by reference herein in its entirety.

[0213] In another example, the immersion cooling system may include one or more pressure relief valves disposed on the tank and / or the bellows. A cross-sectional view of an example pressure relief valve is shown in FIG. 1.28. The pressure relief valve may be a one-way valve that is configured to open when the pressure difference across the valve is greater than or equal to a pressure threshold associated with that valve. The pressure relief valve may open to alleviate undesirable pressure conditions within the tank and / or the bellows. For example, the pressure relief valve may be oriented such that it opens when the tank pressure is greater than the ambient pressure, thus reducing or, in some instances, mitigating overpressure-related failures. In another example, the pressure relief valve may be configured to open when the ambient pressure is greater than the tank pressure, thus reducing or, in some instances, mitigating negative pressure-related failures. Once the tank pressure returns to a desired value, the pressure relief valve may close. As described above, the pressure relief valve(s) may generally be disposed towards the top of the tank and / or the bellows to reduce loss of coolant vapor during an overpressure event.

[0214] The pressure relief valve may support appreciably high flow rates of fluid when open to accommodate the high coolant vapor generation rates associated with high power computing systems. For example, the pressure relief valves disclosed herein may support a flow of a mixture of air and coolant vapor at flow rates up to 500 cfm, 550 cfm, 600 cfm, or even 2000 cfm. Additionally, the pressure threshold at which the pressure relief valve begins to open may be relatively low, for example, between about 0.5 pounds per square inch (psi) and about 2 psi. Additionally, thepressure relief valve may be a passive component, e.g., it does not require active control to open or close the valve.

[0215] As a demonstration, the pressure relief valve was subjected to several tests to evaluate whether the valve opens, closes, and seals as desired. FIG. 1.29 shows the pressure across the valve measured as a function of time. Between 0-4 seconds, the valve was subjected to a ramp up to the cracking pressure (the pressure to open the valve). There was no exhaust once this test was completed, i.e., the experiment moved into fill for the next test. Between 4-29 seconds, the valve was subjected to a ramp down where the pressure decreased to reseat (close) the valve. Between 20-46 seconds, a pressure decay test was preformed to evaluate whether the valve was properly sealed.

[0216] Further details and examples of a pressure relief valve and valve assemblies that include the pressure relief valve may be found in International PCT Application No. PCT / US2023 / 067598, filed on May 30, 2023 and entitled, “PRESSURE RELIEF VALVE AND ADSORBENT CHAMBER FOR TWO-PHASE IMMERSION COOLING SYSTEMS AND METHODS FOR USING SAME,” which is incorporated by reference herein in its entirety.

[0217] When the tank lid is opened, for example, during servicing of the immersion cooling system, coolant vapor may spill out from the tank cavity into the surrounding environment due to the movement of the coolant vapor caused by the boiling of the coolant liquid. This, in turn, may create a safety hazard for personnel operating near the immersion cooling system. Additionally, the loss of coolant vapor and the subsequent replacement of coolant liquid may appreciably increase operating costs.

[0218] The loss of coolant vapor may be reduced, in part, by increasing the height of the tank, which, in turn, increases the head space height within the tank cavity. However, increasing the height of the tank may not be preferable, particularly if the immersion cooling system is deployed in a location with limited space (e.g., the room has a limited height). Thus, another way to reduce coolant vapor loss is by including a diffuser within the head space of the tank cavity, as shown in FIG. 1.30. The diffuser may reduce movement of the coolant vapor at the top of the tank, thus reducing the amount of coolant vapor that spills out from the top opening of the tank.

[0219] To reduce coolant loss through the top opening of the tank when the lid is closed, the lid may include a lid gasket to seal the top opening. In some implementations, the lid may include a single lid gasket that spans the perimeter of the top opening to provide a continuous seal. FIGS. 1.31 A- 1.3 ID show several views of the lid. As shown, the lid may include an outer lid with a lid gasket disposed in a channel formed near the periphery of the outer lid. FIG. 1.31C shows the lidgasket may protrude from the channel. FIG. 1.3 ID shows that when the lid is closed, the lid gasket may physically contact the top plate of the tank and deform to form a seal.

[0220] FIG. 1.32 shows the cross-sectional geometry of the lid gasket. As shown, the lid gasket may include multiple hooks to securely attach the lid gasket to the outer lid. The lid gasket may further include a bulb, which deforms upon contact with the top plate as described above. The lid may be formed from various materials including, but not limited to, silicone.

[0221] FIGS. 1.33A-1.33D show FEA simulation results of the deformation of the lid gasket. In particular, FIG. 1.33B shows that when the lid gasket is compressed, there is no undesirable buckling or interference in the lid gasket. FIGS. 1.33C and 1.33D further show that at the corner portions of the lid gasket undergo significant deformation due to the tight bending radius of the comer. In particular, the results show only 1 mm deformation at the tank lid radii with no buckling.7. 1.7. Coolant Liquid Conditioning

[0222] In practice, the coolant liquids used in the immersion cooling system may have a purity greater than 99.95%. However, the concentration of contaminants in the coolant liquid of a conventional immersion cooling system typically increases over time. The contaminants typically originate from conventional electronic components submerged in the coolant liquid, such as printed circuit boards (PCBs), cables, wires, O-rings, vibration dampers, potting compounds, Formex with fire retardants, and / or the like. Although cooling liquids are poor solvents, they are efficient at extracting materials over time due to constant purification of the fluid by filtration and the distillation of contaminants via boiling. In particular, the solvation of hydrocarbons by the immersion cooling fluid also has the potential to enable co-solvation and mobility of ions and materials otherwise insoluble in the immersion cooling fluid.

[0223] These contaminants can accumulate onto various surfaces of the computing system that dissipate heat (e.g., electrical interconnects, a ball grid array) resulting in performance degradation (e.g., reduction in signal integrity) and, in some instances, premature failure of an electronic component in the computing system. For higher power computing systems, contamination is made worse by the presence of computing components that generate higher heat loads and / or a larger number of heat-generating components.

[0224] FIG. 1.34 provides a summary of the contaminants often contained in the coolant liquid of a two-phase immersion cooling system. These contaminants include, but are not limited to, hydrocarbons and plasticizers (e.g., phthalate esters, aliphatics, and aromatics), UV stabilizers, flame retardants (e.g., tetrabromo(bis)phenol-A (TBBP-A)), and residual solder flux activators. FIGS.1.35A and 1.35B show a cable bill of materials for an example assembled unit to illustrate the number of cables that may be submerged in the coolant liquid during operation. Each of the cables may be formed of materials that can contaminate the coolant liquid.

[0225] To reduce the adverse effects of contamination, the immersion cooling system disclosed herein may include several systems and / or components to condition the coolant liquid (e.g., by reducing the concentration of contaminants). For example, FIG. 1.36 A shows a diagram of the tank, which includes a filtration system to remove contaminants directly from the coolant liquid. The tank may also include one or more desiccant packs to remove water vapor in the head space. During operation, the coolant vapor, air, and water vapor in the head space may become stratified due to their different densities. The coolant vapor may have a higher density than air or water vapor, resulting in a higher concentration of coolant vapor located at the bottom of the head space. Accordingly, the desiccant packs may be situated at the top of the head space where the concentration of water vapor is higher. FIG. 1.36B shows various operating parameters and material requirements for the filtration system and the desiccant.

[0226] FIG. 1.37A shows a diagram of an example filtration system (also referred to herein as an “external filter assembly”). The filtration system may include a filter cartridge containing filtration media (not shown), a pump, hoses, a tank inlet, and a tank outlet. The hoses may be any flexible or non-flexible hoses, pipes, tubes, or ducts. The filtration system may be fluidically coupled with a reservoir of immersion cooling fluid (e.g., the coolant liquid located in the tank cavity).

[0227] FIG. 1.37B shows an example filtration system disposed in a cabinet of the tank. FIG. 1.37C shows a detailed view of the filtration system separate from the tank. As shown, the filtration system may include two filter cartridges - one filter cartridge containing filtration media and the other containing secondary filtration media. The secondary filtration media may be different from the filtration media or may be the same. In some implementations, one or both of the filter cartridges may contain a plurality of filtration media. For example, one filter cartridge may contain activated alumina and activated carbon, while the other filter cartridge may contain a mesh particulate filter.

[0228] Further details and examples of filtration systems and / or systems related to the operation of a filtration system may be found in U.S. Provisional Application No. 63 / 519,090, filed on August 11, 2023 and entitled, “TWO PHASE IMMERSION COOLING FILTRATION MEDIA HEALTH MONITORING,” U.S. Provisional Application No. 63 / 519,080, filed on August 11, 2023 and entitled, “EXTERNAL FILTRATION SYSTEM FOR IMMERSION COOLING WITH THE FLEXIBLE ARRANGEMENT FOR THE MAINTENANCE AND THE APPARATUS THEREOF,” and U.S. Provisional Application No. 63 / 519,088, filed on August 11, 2023 andentitled “FLEXIBLE CAGE-STYLE FILTRATION ASSEMBLY SYSTEM FOR IMMERSION COOLING AND THE COMPONENTS THEREOF.” Each of the aforementioned references is incorporated by reference herein in its entirety.

[0229] FIGS. 1.38A-1.38E show example desiccant assemblies disposed on the top plate of the tank. FIGS. 1.39A-1.39E further show the desiccant assemblies separate from the tank. As shown, the desiccant assemblies may be placed directly below the lid within the head space of the tank cavity. Multiple desiccant assemblies may be disposed across the top opening of the tank to contain desiccant. In some implementations, the multiple desiccant assemblies may collectively store at least 10 kg of silica gel desiccant. For example, the immersion cooling system may include six desiccant assemblies and the 10kg of desiccant may be evenly distributed amongst the desiccant assemblies. In some implementations, each desiccant assembly may be configured to contain at least 2.8 kg of desiccant, thus resulting in a total desiccant capacity of 16.8 kg amongst the six desiccant assemblies. As shown in FIGS. 1.36A-1.36D, each desiccant assembly includes a frame. The frame may function as a tray to carry the desiccant. In some implementations, the frame may include a mounting bracket that rests on the edge of the top plate without any additional coupling mechanism. In this manner, each desiccant assembly may be readily installed by placing the desiccant assembly across the top plate and / or readily removed (e.g., to replace the desiccant after a period of time).

[0230] The desiccant assembly further includes an inner lid disposed onto the frame that covers a portion of the top opening of the top plate. As shown in FIG. 1.38A, when multiple desiccant assemblies are installed, the inner lids of the respective desiccant assemblies may collectively cover the most or, in some instances, all of the top opening. In this manner, when the lid of the tank is opened, the tank cavity may remain enclosed by the inner lids of the desiccant assemblies, thus reducing coolant vapor loss. Additionally, if the desiccant assemblies require servicing (e.g., to replace the desiccant), one desiccant assembly may be removed at a time, thus exposing only a small portion of the top opening to the ambient environment, further reducing coolant vapor loss and / or reducing introduction of water vapor into the tank.

[0231] The desiccant assembly further includes a cage suspended below the frame via multiple tabs (see FIG. 1.39C) and disposed within the head space of the tank cavity. The cage supports a container containing the desiccant. The container may be perforated with multiple openings to allow the gas in the headspace to pass through the desiccant contained in the container. However, the openings may be sufficiently small to reduce or, in some instances, prevent silica gel from falling through the openings and into the coolant liquid located below.

[0232] The cage may be positioned at a distance sufficiently below the frame such that all six sides of the container are exposed to the gas in the head space. In some implementations, the position of the cage and the container with respect to the frame may be adjustable. For example, FIG. 1.39C shows the tabs of the frame include a slot. The cage further includes a pin inserted through the slot, thus allowing the cage and the container to vertically slide along the slot. The cage may be securely coupled to the tab via a nut or a fastener (not shown) to lock the position of the cage and the container along the slot.

[0233] In some implementations, at least half of the cage is located below the top plate of the tank. Preferably, the entirety of the cage is located below the top plate. In some implementations, the container may have a height of about 60 mm. Additionally, the slot on the tabs may have a length of about 40 mm, thus providing nominally 40 mm of adjustment to the vertical position of the cage and the container. When the cage is at its lowest position along the tabs, the horizontal centerline of the container may be located about 60 mm below the interior surface of the lid of the tank. The bottom side of the container may be disposed above the condenser conduits by about 40 mm.

[0234] FIGS. 1.39D and 1.39E show the container may include a container cover and a container base. The container cover may be readily placed onto the container base and / or removed from the container base to facilitate removal of used desiccant and refill of new desiccant. A latch may be rotatably coupled to the container cover via a pivot pin. The container base may further include a locking pin. When the container cover is placed onto the container base, the container cover may be locked to container base via the latch and the locking pin. The cage may further include handles disposed on opposing sides to improve ease of handling when positioning the cage with respect to the frame.

[0235] The inner lid may be formed from transparent materials including, but not limited to, polycarbonate and acrylic. The frame, the cage, and the container may be formed from various metals including, not limited to, stainless steel.

[0236] The concentration of contaminants in the coolant liquid may also be reduced through careful selection of materials that are compatible with the coolant liquid (e.g., materials that do not readily leach out contaminants) and / or modifying sensitive components that are particularly susceptible to contamination (e.g., heat-generating electronic components / devices).

[0237] For example, underfill or edge bonds may be applied to any heat generating electronic device (e.g., a ball grid array (BGA) device). Further examples of mitigation strategies for reducing contamination using an edge bond, underfill, or a protective coating may be found in U.S. Provisional Application No. 63 / 503,401, filed on May 19, 2023 and entitled, “Edge Bonding forImmersion Systems,” U.S. Provisional Application No. 63 / 503,405, filed on May 19, 2023 and entitled, “Application of Underfilling BGA Devices in Two Phase Immersion Cooling,” U.S. Provisional Application No. 63 / 505,129, filed on May 31, 2023 and entitled, “PROTECTIVE COATING FOR ELECTRONIC DEVICES IN IMMERSION COOLING SYSTEMS,” and U.S. Provisional Application No. 63 / 514,103, filed on July 17, 2023 and entitled, “EDGE SEALED ELECTRONIC DEVICE FOR IMMERSION COOLING ENVIRONMENT AND METHODS FOR MAKING SAME.” Each of the aforementioned references is incorporated by reference herein in its entirety.

[0238] In another example, the immersion cooling system may preferably use a busbar for power distribution rather than cables. In yet another example, it may be preferable to blind-mate back plane power and management connections instead of cables. In yet another example, thermal gap pads may be eliminated. In yet another example, potting compounds and vibration dampers may be removed. In yet another example, the immersion cooling system may use relatively fewer passthroughs and / or seals / gaskets. In yet another example, it is preferable to use materials that are free from additives such as flame retardants and UV stabilizers.

[0239] If cables are included in the immersion cooling system, it is preferable to include cables with insulation materials formed of fluorinated polymers and / or cross-linked polyolefins. Plasticized PVC and polyurethanes should not be used. Additionally, highly cross-linked thermal pastes are preferable over conventional thermal interface material.

[0240] Additionally, solder flux residue presents a high risk for contamination in two phase immersion cooling. Thus, PCB cleaning should be well-controlled. In some implementations, it the tank filtration system may be run while computing system is powered down to ‘pre-clean’ materials in the tank.

[0241] The filtration system described above may operate continuously to ensure fluid purity over the lifetime of the equipment. As described above, the filters may utilize activated carbon and / or alumina to remove hydrocarbon contaminants and other impurities from the fluid. Additionally, sensors disposed in the coolant liquid may provide feedback on filtration system performance and overall tank health by directly measuring contamination concentration (see examples of filtration systems described above and incorporated by reference).8. 1.8. Safety Standards & Servicing

[0242] The immersion cooling systems disclosed herein may meet various safety and environmental standards set forth by various organizations including, but not limited to, the InternationalElectrotechnical Commission (IEC), and the European Union (EU). For example, the immersion cooling systems disclosed herein may meet the requirements defined by the IEC standard IEC 62368-1 entitled, “Audio / video, information and communication technology equipment - Part 1: Safety requirements,” which standard is hereby incorporated herein by reference. In another example, the immersion cooling systems disclosed herein may meet the requirements defined by the EU Restriction on Hazardous Substances (RoHS) directive, which standard is hereby incorporated herein by reference.

[0243] The immersion cooling system may generally include components that may be serviced and / or replaced overtime. These components include, but is not limited to, the filter pump for the filtration system, the carbon filter for the filtration system, bellows, the sensor bar, the desiccant assembly, and the pressure transducers of the filtration system. FIGS. 1.40A and 1.40B provide example procedures for servicing and / or replacing these foregoing components.9. 1.9. Manufacture, Assembly, & Validation

[0244] During manufacture and assembly of the immersion cooling system, various tests may be performed to validate different subassemblies (e.g., the bellows, bus bar passthrough, the pressure relief valve, the seals, the condenser, and the fiber optic passthrough), then the tank (e.g., the tank itself, the tank with the various subassemblies mounted thereto), and then the fully assembled immersion cooling system. For example, various leak tests may be performed during this process (e.g., from the subassemblies to the tank and then the immersion cooling system). FIG. 1.41 A provides a table of some of the validation tests that may be performed on various subassemblies of the immersion cooling system. FIG. 1.41B provides a table of some of validation tests that may be performed on the assembled tank and / or the immersion cooling system.

[0245] The immersion cooling systems disclosed herein may operate using relatively high purity coolant liquids. For example, the purity may be greater than 99%, more preferably greater than 99.5%, and even more preferably greater than 99.9%. In one example, the coolant liquid may be a 3M Novec 649 Engineered Fluid. In another example, FIG. 1.42 shows the specifications for an example coolant liquid. The specifications may be based on industry standards related to fire extinguishing, such as the National Fire Protection Association (NFPA) 2001 standard entitled, “Standard on Clean Agent Fire Extinguishing Systems,” which standard is hereby incorporated herein by reference. The coolant liquid may be provided by various suppliers, such as 3M, Inventec, Envirotech, and Unistar.10. 1.10. Server Lift System

[0246] In some implementations, it may be desirable to quickly install and / or remove the components of the computing system (e.g., the assembled units) to / from the immersion cooling system to facilitate rapid deployment, servicing, and / or testing. For example, newly manufactured assembled units may be loaded into a test tank, tested at normal or accelerated operation, and unloaded from the test tank to make way for another batch of newly manufactured assembled units. Thus, rapid installation and removal of the assembled units may increase throughput. In another example, assembled units may be removed and replaced during operation. In this example, rapid installation and removal of the assembled units may reduce system downtime.

[0247] However, in some implementations, the components of the computing system may be relatively large and heavy (e.g., weighing at least 150 pounds or more), thus making installation slower and more challenging. Additionally, the immersion cooling systems disclosed herein may include several assembled units (e.g., 32 assembled units, 64 assembled units) as described above. Furthermore, different assembled units may be installed in an immersion cooling system. For example, FIG. 1.43 shows the tank of FIG. 1.2A with assembled units A and B installed in the tank cavity.

[0248] FIG. 1.44 shows magnified view of the top portion of assembled unit A. As shown, the assembled unit A includes a housing to contain various electronic components and a mounting tab mounted to the top of the housing. The mounting tab may include a lift opening to facilitate insertion and / or removal. FIGS. 1.45A and 1.45B show several views of the assembled unit B. As shown, the assembled unit B may include a housing with a mounting tab and a lift opening to facilitate insertion and / or removal. The assembled unit B may further include multiple gripper receptacles to provide another mechanism to facilitate insertion and / or removal.

[0249] To rapidly install and / or remove assembled units to / from the immersion cooling system, a server lift system may be deployed together with the immersion cooling system. The server lift system may be used during manufacture to validate the assembled units and / or at the facility to facilitate installation and / or removal of assembled units for deployment.

[0250] The server lift system may include, for example, a gantry system and one or more end effectors to facilitate placement and / or retrieval of an assembled unit. FIG. 1.46 shows an example gantry system that includes a pair of rails and a crane that slidably moves along the rails. In some implementations, the gantry system may be a PWI UltraLite Freestanding Bridge Crane. FIG. 1.47A shows an example end effector that may be mounted to the crane of the gantry system. The end effector may include a pair of grippers. The end effector may further include a gripper motorto actuate the grippers, a guide attach to guide the placement and positioning of the end effector with respect to the tank, and a mounting hook for attached to the crane. In some implementations, the gripper motor may be a Zimmer GPP5016N-2, as shown in FIG.l. 47B.

[0251] The pair of grippers may be configured to engage the gripper receptacles of the assembled unit B. For example, FIG. 1.48 shows the grippers may first be inserted into respective gripper receptacles on the assembled unit B. The grippers may be displaced by the gripper motor to engage the gripper receptacles. Thereafter, the end effector may be raised to lift the assembled unit B.

[0252] It should be appreciated that the pair of grippers is a non-limiting example. The end effector may, alternatively, include a hook or a latch mechanism to attach to the lift openings on the assembled units A and B. In some implementations, the lift system may include multiple, different end effectors to accommodate different mounting points for different assembled units.

[0253] FIG. 1.49 shows a top view of a server lift system deployed together with a pair of immersion cooling systems. As shown, the server lift system may be configured to service both immersion cooling systems. For example, the crane may move overhead of either immersion cooling systems and the end effector attached to the crane may be lowered to insert or remove an assembled unit accordingly.

[0254] Further examples of a lift system to facilitate installation and / or removal of an assembled unit to / from the tank may be found in U.S. Provisional Application No. 63 / 519,075, filed on August 11, 2023 and entitled, “Testing of Immersion-Cooled Servers in a Manufacturing Environment,” which is incorporated by reference herein in its entirety.11. 1.11. Conclusion

[0255] All parameters, dimensions, materials, and configurations described herein are meant to be example and the actual parameters, dimensions, materials, and / or configurations will depend upon the specific application or applications for which the inventive teachings is / are used. It is to be understood that the foregoing embodiments are presented primarily by way of example and that, within the scope of the appended claims and equivalents thereto, inventive embodiments may be practiced otherwise than as specifically described and claimed. Inventive embodiments of the present disclosure are directed to each individual feature, system, article, material, kit, and / or method described herein.

[0256] In addition, any combination of two or more such features, systems, articles, materials, kits, and / or methods, if such features, systems, articles, materials, kits, and / or methods are not mutually inconsistent, is included within the inventive scope of the present disclosure. Othersubstitutions, modifications, changes, and omissions may be made in the design, operating conditions and arrangement of respective elements of the example implementations without departing from the scope of the present disclosure. The use of a numerical range does not preclude equivalents that fall outside the range that fulfill the same function, in the same way, to produce the same result.

[0257] Also, various inventive concepts may be embodied as one or more methods, of which an example has been provided. The acts performed as part of the method may be ordered in any suitable way. Accordingly, embodiments may be constructed in which acts are performed in an order different than illustrated, which may include performing some acts simultaneously, even though shown as sequential acts in illustrative embodiments.

[0258] All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety.

[0259] All definitions, as defined and used herein, should be understood to control over dictionary definitions, definitions in documents incorporated by reference, and / or ordinary meanings of the defined terms.

[0260] The indefinite articles “a” and “an,” as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.”

[0261] The phrase “and / or,” as used herein in the specification and in the claims, should be understood to mean “either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with “and / or” should be construed in the same fashion, i.e., “one or more” of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by the “and / or” clause, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, a reference to “A and / or B”, when used in conjunction with open-ended language such as “comprising” can refer, in one embodiment, to A only (optionally including elements other than B); in another embodiment, to B only (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0262] As used herein in the specification and in the claims, “or” should be understood to have the same meaning as “and / or” as defined above. For example, when separating items in a list, “or” or “and / or” shall be interpreted as being inclusive, i.e., the inclusion of at least one, but also including more than one, of a number or list of elements, and, optionally, additional unlisted items. Only terms clearly indicated to the contrary, such as “only one of’ or “exactly one of,” or, when used in the claims, “consisting of,” will refer to the inclusion of exactly one element of a numberor list of elements. In general, the term “or” as used herein shall only be interpreted as indicating exclusive alternatives (i.e., “one or the other but not both”) when preceded by terms of exclusivity, such as “either,” “one of,” “only one of,” or “exactly one of.” “Consisting essentially of,” when used in the claims, shall have its ordinary meaning as used in the field of patent law.[0263| As used herein in the specification and in the claims, the phrase “at least one,” in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase “at least one” refers, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, “at least one of A and B” (or, equivalently, “at least one of A or B,” or, equivalently “at least one of A and / or B”) can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.

[0264] In the claims, as well as in the specification above, all transitional phrases such as “comprising,” “including,” “carrying,” “having,” “containing,” “involving,” “holding,” “composed of,” and the like are to be understood to be open-ended, i.e., to mean including but not limited to. Only the transitional phrases “consisting of’ and “consisting essentially of’ shall be closed or semiclosed transitional phrases, respectively, as set forth in the United States Patent Office Manual of Patent Examining Procedures, Section 2111.03.

[0265] Section 2: Multi-Purpose Sleeve for Immersion-Cooled Servers A typical immersion tank for immersion cooling of servers and other electronic components is a bulky, welded mechanical assembly with large dimensional tolerances. For instance, the slots in the immersion tank for servers may be from 8 mm larger to 5 mm smaller than intended due to manufacturing imperfections. When the immersion tank is filled with immersion fluid and pressurized, the immersion tank’s walls may expand and / or contract, causing the slots for the servers to expand and / or contract as well, e.g., by up to about ± 8 mm (± 3 mm).

[0266] To account for the large manufacturing tolerances and to avoid interference with the servers , the immersion tank should be sized larger than the servers . For instance, if a server’ s maximum width is 500 mm, then the slot in the immersion tank for that server should have a nominal widthof at least 513 mm to account for the potential variation of the manufacturing tolerance of -5 mm, plus an extra -8 mm to account for immersion tank wall movement due to negative pressure condition, resulting in larger immersion tanks and larger slots for servers in the immersion tanks. When the immersion tank is on the larger side of the tolerance, the maximum immersion tank width could be as high as 521 mm (unpressurized) or 529 mm (pressurized). These manufacturing tolerances also result in big variations in server openings from tank to tank.

[0267] The larger sizes and size variations among openings for servers in immersion tanks create several problems. For instance, when the slot or opening in the immersion tank is larger than the server, then aligning the server to the immersion tank can be more difficult as the server can be biased on one side of the tank and may not connect or couple properly to the bus bar inside the immersion tank. When the slot or opening is smaller, on the other hand, the server chassis can rub against the guide rails instead the immersion tank as the server is being inserted into or removed from the immersion tank. The immersion tank and server chassis are typically made of steel or another metal, so this rubbing can produce metal debris. As the immersion fluid boils, it moves from bottom of the tank to top, bringing the metal debris into contact with the servers and / or other electrical components, which may cause electrical shorts. Rubbing between metal surfaces can also produce metal debris even for larger slots and openings. The larger slots for servers also lead to larger immersion tank volumes, which in turn leads to larger amounts of immersion fluid needed to fill the immersion tanks. Immersion fluid is very expensive, so the extra immersion fluid increases cost without necessarily improving cooling performance.|0268] FIG. 2.1 shows a gasket or sleeve 100 that addresses problems associated with large manufacturing tolerances for slots or openings in immersion tanks. The sleeve 100 allows for “less- than-perfect” mating of the server chassis to guide rails and / or alignment features within the immersion tanks. The sleeve 100 can be attached to the side of an immersion-cooled server chassis before the server chassis is inserted into a slot or opening in the immersion tank. The server chassis may have a height of 700 mm or 925 mm, a length of 797 mm, and a thickness of 86.9 mm, and the sleeve 100 may have a length of about 700 mm or 925 mm (to match the height of the server chassis), a width of about 85 mm (to roughly match the thickness of the server chassis), and a thickness of about 8 mm (to fill the expected maximum gap between the server rails and the immersion tank). The sleeve 100 prevents metal-on-metal contact between the server chassis and the guide rails within the immersion tank, reducing the probability of creating metal debris when inserting or removing the server chassis. And the sleeve 100 occupies space inside the immersion tank that would otherwise be occupied by the immersion fluid, reducing the quantity of immersion fluid needed to fill the immersion tank.

[0269] The sleeve 100 is made of a strip of electrically insulating, elastically deformable material, such as rubber, silicone, plastic, or polytetrafluoroethylene (PTFE). For instance, the sleeve 100 can be stamped out of a single sheet or material, molded, or extruded. This material should also be inert; that is, it should not react with the immersion fluid, anything else within the immersion tank, or the immersion tank itself. In addition, the material should not degrade or decompose in the immersion tank or when it is inserted into or removed from the immersion tank because degradation could affect the immersion fluid’s cooling capacity. Degradation could also deleteriously affect the operation of the server and / or other components within the immersion tank.

[0270] The sleeve 100 includes two or more snap-fit features 110 or other fasteners for attaching the sleeve 100 to a server chassis. In FIG. 2.1 , for example, the snap-fit features 110 are at the ends of the sleeve 100 and secure the sleeve 100 to a server chassis in a detachable fashion. The snap- fit features 110 in FIG. 2.1 have mushroom heads that fit through holes in the server chassis and grab the inside wall of the server chassis (e.g., as shown in FIG. 2.2A, described below). Other sleeves may have different types of fasteners, such as snaps, buttons, latches, and so on, that allow compression and decompression of the sleeve. They may also have more fasteners, for example, disposed along the length of the sleeve 100. Alternatively, or in addition, the sleeve 100 can be shaped to conform to or stretch or wrap around the server chassis such that the sleeve 100 stays secured to the server chassis.

[0271] There are also one or more springs or spring features 120 disposed along the length of the sleeve 100. These spring features 120 protrude from the sleeve and can expand or contract to adjust to a wide variety of immersion tank slot sizes resulting from the large manufacturing tolerances for immersion tanks. When the sleeve 100 is attached to the server chassis and the server chassis is in the immersion tank, the spring features 110 on the sleeves 100 push outwards, holding the sleeves 100 and the server chassis in place with respect to the immersion tank. When the spring features 120 are fully expanded against the immersion tank, the combined width of the server chassis and the sleeve(s) 100 should equal the maximum width of the opening or slot in the immersion tank. This means that when the spring features 120 are in a free or fully expanded state, the sleeve 100 at its thickest point should be wide enough to fill the largest possible gap between the server chassis and the rails that define the opening or slot in the immersion tank for the server chassis. Because the rails take up some space, the maximum spring thickness may be about 7 or 8 mm, for example, and may compress by about 7 mm.

[0272] The sleeve 100 also has chamfered or round ends 130 to aid in alignment and to reduce the possibility of contact between the metal server chassis and the metal guide rails of the immersion tank. Preventing or reducing metal-to-metal contact prevents or reduces the possibility of creatingmetal debris that could cause electrical shorting of servers and other components in the immersion tank. FIG. 2.1 shows that both ends of the sleeve 100 are chamfered (cut away to form a sloping edge) or rounded. Alternatively, the sleeve may have a single chamfered end — the end that fits around the comer of the server chassis that is inserted first into the immersion tank.[0273| FIGS. 2.2A-2.2C show different views of immersion sleeves 100 installed on both sides of a server chassis 200 that holds a server 202. FIG. 2.2A shows a front view of the server chassis 200, illustrating the server 202 and one of the guide pins 210 that extends from the server chassis 200 for aligning the server chassis 200 to the immersion tank. FIG. 2.2A also shows the mushroom head of one fastener 110 sticking through a hole in the server chassis 200 and one of the sleeve’s chamfered ends 130. The sleeve’s spring features 120 extend towards the server chassis 200 as shown in FIGS. 2.2A and 2.2C. Alternative sleeves may have spring features that extend away from the server chassis 200 when installed on the server chassis 200. FIGS. 2.2B and 2.2C show that the sleeves 100 are as wide as the server chassis 200. The sleeves 100 can be shorter than the server chassis 200 is long, as long as the server chassis 200, or even longer than the server chassis 200, with the extra length extending towards the top of immersion tank and possibly forming a handle for removing the server chassis 200 from the immersion tank.

[0274] FIGS. 2.3A-2.3D illustrate the server chassis 200 with sleeves 100 along both sides installed in an immersion tank 300. FIG. 2.3A shows a vertical cross-section of the immersion tank 300, and FIGS. 2.3B-2.3D show top-down views of the immersion tank 300 at different levels of detail. The immersion tank 300 includes guide rails 302 and an alignment plate 310 that align the server chassis 200 to the immersion tank 300 and hold the server chassis 200 in position. The alignment plate 310 and a bus bar 320 sit at the bottom of the immersion tank 300 between condenser coils 330 that cool vaporized immersion fluid.

[0275] The server chassis 200 can come in different sizes, e.g., with heights of 700 mm or 925 mm. When fully inserted into the immersion tank 300, the shorter server chassis 200 may be below the condenser coils 330 and may be fully immersion in the immersion fluid (not shown). The taller server chassis 200 may not fit completely below the condenser coils 330 and may be partially immersed in the immersion fluid. The guide pins 210 extending from the server chassis 200 fit into guide pin receptacles 312 in the alignment plate 310. Electrical pins 204 electrically coupled to the server 202 and extending from the server chassis 200 contact bus bar connections 314 that are electrically coupled to the bus bar 320, which provides electrical power to the server 202.

[0276] When the server chassis 200 is installed in the immersion tank 300, the chamfer 130 on the sleeve 100 helps to align the server chassis 200 to the guide rails 302 in the immersion tank 300. In the cases where the tank opening is smaller than the total width of the server chassis 200 andsleeves 100, as the server chassis 200 slides down into the immersion tank 300, the guide rails 302 push the sleeve 100 toward the server chassis 200, compressing the sleeves’ spring features 120. When the server chassis 200 is pulled out of the immersion tank 300, the spring features 120 expand back to their original state. This expansion and contraction of the sleeve’s spring features 120 account for the large tank manufacturing tolerances (e.g., +8 / -5 mm (static) and ±8 mm (pressurized)).

[0277] FIG. 2.4 shows top, side, and cross-sectional views of an alternative gasket or sleeve 400 for an immersion-cooled server chassis. This sleeve 400 is a tube-like structure that is made of compliant, elastically deformable, electrically insulating material, such as rubber, plastic, or PTFE. It has two or more fasteners 410 for attaching the sleeve 400 to the server chassis and rounded or chamfered ends 430 for sliding the server chassis into the immersion tank. The tube-like structure is hollow and can collapse, e.g., along predefined fold lines or pleats, to accommodate narrower openings as shown in the cross-sectional views. The tube-like structure can have a square or rectangular cross section as shown in FIG. 2.4 or a curved or pleated cross section. A vent or hole 420 in the sleeve 400 allows fluid (e.g., air or immersion fluid) to enter and exit the sleeve 400 as the sleeve 400 expands or is compressed. Placing the vent 420 at the end of the sleeve 400 that is not immersed in immersion fluid as shown in FIG. 2.4 allows air or vapor to enter the sleeve 400 and reduces the amount of immersion fluid needed to fill the immersion tank. (If the server chassis is short enough to be completely immersed in the immersion fluid, the sleeve 400 may be made long enough so that one end sticks out of the immersion fluid when the server chassis and sleeve 400 are installed in the immersion tank.) Alternatively, the sleeve 400 can have one or more vents or holes along its length. Or instead of being hollow and having vents or holes, the sleeve 400 can be made of a sponge-like material that can be compressed and springs backs into shape when relaxed.

[0278] While various inventive embodiments have been described and illustrated herein, those of ordinary skill in the art will readily envision a variety of other means and / or structures for performing the function and / or obtaining the results and / or one or more of the advantages described herein, and each of such variations and / or modifications is deemed to be within the scope of the inventive embodiments described herein. More generally, those skilled in the art will readily appreciate that all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and / or configurations will depend upon the specific application or applications for which the inventive teachings is / are used. Those skilled in the art will recognize or be able to ascertain, using no more than routine experimentation, many equivalents to the specific inventive embodiments described herein. It is,therefore, to be understood that the foregoing embodiments are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, inventive embodiments may be practiced otherwise than as specifically described and claimed. Inventive embodiments of the present disclosure are directed to each individual feature, system, article, material, kit, and / or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and / or methods, if such features, systems, articles, materials, kits, and / or methods are not mutually inconsistent, is included within the inventive scope of the present disclosure.

[0279] Also, various inventive concepts may be embodied as one or more methods, of which an example has been provided. The acts performed as part of the method may be ordered in any suitable way. Accordingly, embodiments may be constructed in which acts are performed in an order different than illustrated, which may include performing some acts simultaneously, even though shown as sequential acts in illustrative embodiments.

[0280] All definitions, as defined and used herein, should be understood to control over dictionary definitions, definitions in documents incorporated by reference, and / or ordinary meanings of the defined terms.

[0281] The indefinite articles “a” and “an,” as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.”

[0282] The phrase “and / or,” as used herein in the specification and in the claims, should be understood to mean “either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with “and / or” should be construed in the same fashion, i.e., “one or more” of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by the “and / or” clause, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, a reference to “A and / or B”, when used in conjunction with open-ended language such as “comprising” can refer, in one embodiment, to A only (optionally including elements other than B); in another embodiment, to B only (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0283] As used herein in the specification and in the claims, “or” should be understood to have the same meaning as “and / or” as defined above. For example, when separating items in a list, “or” or “and / or” shall be interpreted as being inclusive, i.e., the inclusion of at least one, but also including more than one, of a number or list of elements, and, optionally, additional unlisted items. Only terms clearly indicated to the contrary, such as “only one of’ or “exactly one of,” or, whenused in the claims, “consisting of,” will refer to the inclusion of exactly one element of a number or list of elements. In general, the term “or” as used herein shall only be interpreted as indicating exclusive alternatives (i.e., “one or the other but not both”) when preceded by terms of exclusivity, such as “either,” “one of,” “only one of,” or “exactly one of.” “Consisting essentially of,” when used in the claims, shall have its ordinary meaning as used in the field of patent law.

[0284] As used herein in the specification and in the claims, the phrase “at least one,” in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase “at least one” refers, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, “at least one of A and B” (or, equivalently, “at least one of A or B,” or, equivalently “at least one of A and / or B”) can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.

[0285] In the claims, as well as in the specification above, all transitional phrases such as “comprising,” “including,” “carrying,” “having,” “containing,” “involving,” “holding,” “composed of,” and the like are to be understood to be open-ended, i.e., to mean including but not limited to. Only the transitional phrases “consisting of” and “consisting essentially of’ shall be closed or semiclosed transitional phrases, respectively, as set forth in the United States Patent Office Manual of Patent Examining Procedures, Section 2111.03.

[0286] A two-phase immersion cooling system can be summarized as including: a tank having a tank cavity, wherein the tank is configured to contain a coolant liquid and a computing system submerged in the coolant liquid; bellows coupled to a top plate of the tank; one or more pressure relief valves disposed on the bellows and tank; a pair of power racks coupled to the tank by at least one busbar; and a pair of grounding assemblies disposed on outer walls of the tank.

[0287] The coolant liquid and the computing system are disposed in a lower portion of the tank cavity defined by the tank. The tank includes a lid configured to open to access the tank cavity and configured to close to seal the tank cavity. The tank includes a pair of condensers partly disposed in an upper portion of the tank cavity. The condensers are configured to extract heat from coolantvapor generated from the coolant liquid during operation, causing the coolant vapor to condense back to a liquid and fall into a pool of coolant liquid located in the lower portion of the tank cavity. Each respective condenser is configured to extract heat from the coolant vapor using a fluid circulated by an external fluidic system coupled to the tank. Each respective condenser includes one water box fluidically coupled to an inlet pipe to receive water at a first temperature from an external water source, and another water box coupled to an outlet pipe to discharge water at a second temperature to the external water source. The pair of water boxes are fluidically coupled together by a plurality of fluid conduits disposed within the tank cavity.

[0288] The bellows are coupled to the top plate of the tank by multiple venting conduits coupled to respective ports of the tank and the bellows. The bellows are configured to regulate tank pressure during operation by providing an additional space to store gas from the tank when tank pressure increases and to supply gas to the tank when tank pressure decreases. Each respective pressure relief valve is configured to open when pressure difference across the pressure relief valve exceeds a pressure threshold associated with the respective pressure relief valve. Each respective pressure relief valve is configured to open only when tank pressure is greater than ambient pressure. Each respective pressure relief valve is configured to open only when ambient pressure is greater than the tank pressure.

[0289] The pair of power racks are separate from the tank and configured to supply electrical power to the computing system disposed within the tank. Each respective grounding assembly includes a ground connection for the tank, and wherein the grounding assembly includes one or more ground cables coupled to a bracket mounted to a tank wall of the tank.

[0290] A system can be summarized as including: a tank including a tank frame supporting a plurality of tank walls that define a cavity; a top plate of the tank defining a top opening; a computing system inserted through the top opening and into the cavity; and a busbar coupled to multiple units of the computing system.

[0291] The system further includes a lid on the tank configured to move between a first position and a second position to provide access to the cavity. The lid is closed via a locking mechanism. The locking mechanism includes a spring-loaded latch mechanism. The cavity is divided into a lower portion and an upper portion, wherein the lower portion includes the computing system and a coolant liquid, and the upper portion includes a pair of condensers. The cavity is configured to contain coolant liquid at a first height defined from a bottom of the cavity to a first surface of a coolant liquid of at least 0.7 m. Each respective condenser includes a pair of water boxes mounted to an exterior portion of the tank.

[0292] Multiple ports disposed on the top plate fluidically couples the tank to bellows via venting conduits. Multiple busbar feedthroughs facilitate insertion of a busbar from a power rack into the cavity. The system further includes one or more cabinets having auxiliary systems, wherein the auxiliary systems include control electronics and a filtration system.

[0293] A gasket for a server chassis can be summarized as including: an elastically deformable material formed from the gasket and having at least one fastener for securing the gasket to the server chassis. The gasket is configured to prevent the server chassis from rubbing against a metal surface of an immersion tank as the server chassis and the gasket are inserted into the immersion tank. The gasket is configured to be elastically compressed. The gasket is made of at least one of plastic, rubber, silicone, plastic, or polytetrafluoroethylene.

[0294] The gasket further includes at least one compressible spring feature configured to be compressed between the server chassis and an immersion tank. The gasket includes a sponge-like material. The gasket defines a hollow, compressible lumen having a vent or hole to allow fluid flow into and out of the hollow, compressible lumen. The vent or hole is disposed at an end of the gasket. The gasket also has a chamfered corner.

Claims

CLAIMS1. A two-phase immersion cooling system, comprising: a tank having a tank cavity, wherein the tank is configured to contain a coolant liquid and a computing system submerged in the coolant liquid; bellows coupled to a top plate of the tank; one or more pressure relief valves disposed on the bellows and tank; a pair of power racks coupled to the tank by at least one busbar; and a pair of grounding assemblies disposed on outer walls of the tank.

2. The cooling system of claim 1, wherein the coolant liquid and the computing system are disposed in a lower portion of the tank cavity defined by the tank.

3. The cooling system of claim 1, wherein the tank comprises a lid configured to open to access the tank cavity and configured to close to seal the tank cavity.

4. The cooling system of claim 1, wherein the tank includes a pair of condensers partly disposed in an upper portion of the tank cavity.

5. The cooling system of claim 4, wherein the condensers are configured to extract heat from coolant vapor generated from the coolant liquid during operation, causing the coolant vapor to condense back to a liquid and fall into a pool of coolant liquid located in the lower portion of the tank cavity.

6. The cooling system of claim 5, wherein each respective condenser is configured to extract heat from the coolant vapor using a fluid circulated by an external fluidic system coupled to the tank.

7. The cooling system of claim 6, wherein each respective condenser includes one water box fluidically coupled to an inlet pipe to receive water at a first temperature from an external water source, and another water box coupled to an outlet pipe to discharge water at a second temperature to the external water source.

8. The cooling system of claim 7, wherein the pair of water boxes are fluidically coupled together by a plurality of fluid conduits disposed within the tank cavity.

9. The cooling system of claim 1, wherein the bellows are coupled to the top plate of the tank by multiple venting conduits coupled to respective ports of the tank and the bellows.

10. The cooling system of claim 1, wherein the bellows are configured to regulate tank pressure during operation by providing an additional space to store gas from the tank when tank pressure increases and to supply gas to the tank when tank pressure decreases.

11. The cooling system of claim 1, wherein each respective pressure relief valve is configured to open when pressure difference across the pressure relief valve exceeds a pressure threshold associated with the respective pressure relief valve.

12. The cooling system of claim 1, wherein each respective pressure relief valve is configured to open only when tank pressure is greater than ambient pressure.

13. The cooling system of claim 1, wherein each respective pressure relief valve is configured to open only when ambient pressure is greater than the tank pressure.

14. The cooling system of claim 1, wherein the pair of power racks are separate from the tank and configured to supply electrical power to the computing system disposed within the tank.

15. The cooling system of claim 1, wherein each respective grounding assembly includes a ground connection for the tank, and wherein the grounding assembly includes one or more ground cables coupled to a bracket mounted to a tank wall of the tank.

16. A system, comprising: a tank including a tank frame supporting a plurality of tank walls that define a cavity; a top plate of the tank defining a top opening; a computing system inserted through the top opening and into the cavity; and a busbar coupled to multiple units of the computing system.

17. The system of claim 16, further comprising a lid on the tank configured to move between a first position and a second position to provide access to the cavity.

18. The system of claim 17, wherein the lid is closed via a locking mechanism.

19. The system of claim 18, wherein the locking mechanism includes a spring-loaded latch mechanism.

20. The system of claim 16, wherein the cavity is divided into a lower portion and an upper portion, wherein the lower portion includes the computing system and a coolant liquid, and the upper portion includes a pair of condensers.

21. The system of claim 16, wherein the cavity is configured to contain coolant liquid at a first height defined from a bottom of the cavity to a first surface of a coolant liquid of at least 0.7 m.

22. The system of claim 20, wherein each respective condenser includes a pair of water boxes mounted to an exterior portion of the tank.

23. The system of claim 16, wherein multiple ports disposed on the top plate fluidi- cally couples the tank to bellows via venting conduits.

24. The system of claim 16, wherein multiple busbar feedthroughs facilitate insertion of a busbar from a power rack into the cavity.

25. The system of claim 16, further comprising one or more cabinets having auxiliary systems, wherein the auxiliary systems include control electronics and a filtration system.

26. A gasket for a server chassis, comprising: an elastically deformable material formed from the gasket and having at least one fastener for securing the gasket to the server chassis.

27. The gasket of claim 26, wherein the gasket is configured to prevent the server chassis from rubbing against a metal surface of an immersion tank as the server chassis and the gasket are inserted into the immersion tank.

28. The gasket of claim 26, wherein the gasket is configured to be elastically compressed.

29. The gasket of claim 26, wherein the gasket is made of at least one of plastic, rubber, silicone, plastic, or polytetrafluoroethylene.

30. The gasket of claim 26, further comprising at least one compressible spring feature configured to be compressed between the server chassis and an immersion tank.

31. The gasket of claim 26, wherein the gasket comprises a sponge-like material.

32. The gasket of claim 26, wherein the gasket defines a hollow, compressible lumen having a vent or hole to allow fluid flow into and out of the hollow, compressible lumen.

33. The gasket of claim 32, wherein the vent or hole is disposed at an end of the gasket.

34. The gasket of claim 26, wherein the gasket has a chamfered corner.

Images