Control of cooling system for heat reuse in data centers

Abstract

A system and method for controlling heat reuse in a data center is provided. The system includes a plurality of evaporator units (EUs) each configured to operate as a two-phase evaporator, each EU attached to a central processing unit (processing unit) package of a data center server and drawing heat from the processing unit package. The system further includes a heat reuse unit (HRU) condenser configured to supply refrigerant liquid to the plurality of EUs and receive heated refrigerant from the plurality of EUs, and further configured to dump heat from the heated refrigerant into an HRU media that supplies heat to a reuse application, and an HRU controller including programmed instructions that, when executed by the HRU controller, perform steps to gradually reduce the flow rate of the HRU media such that temperatures of all processing units are gradually increased in a balanced manner to maximize refrigerant temperatures for heat reuse.

Description

[Technical field]

[0001] The present invention relates generally to the field of room cooling, and more particularly to data center cooling and heat reuse. [Background technology]

[0002] The demand for data processing, data storage, and telecommunications has skyrocketed over the past few decades, resulting in a dramatic growth in the data center industry. Data centers are a major consumer of energy worldwide, consuming an estimated 2% of all electricity generation.

[0003] Given the high cost of energy and the environmental impact of generating electricity, data center owners continue to look for ways to reduce the energy requirements of their facilities. Manufacturers of data processing equipment have aided in this effort by continuing to look for ways to reduce the energy required for a given data processing load. The energy efficiency of data centers can also be improved by reusing the heat they generate. Heat from data centers can be reused to heat both the air and water in residential and commercial buildings. Additionally, many projects around the world use "waste" heat for district heating.

[0004] Most air conditioning systems that extract heat from data centers provide it as relatively cool air or water, reducing the potential efficiency benefits: Typically, the heat must be further processed by a heat pump to be effective, and heated air is a highly inefficient way to transfer heat over long distances.

[0005] Therefore, methods are needed to improve the "quality" of waste heat to facilitate more efficient reuse. Summary of the Invention [Problem to be solved by the invention]

[0006] SUMMARY OF THE DISCLOSURE Embodiments of the present invention provide a system and method for control of a cooling system for heat reuse in a data center. [Means for solving the problem]

[0007] An embodiment may include a plurality of evaporator units (EUs), each configured to operate as a two-phase evaporator, each affixed to a respective processing unit package of a data center server and drawing heat from the respective processing unit package, a heat reuse unit (HRU) condenser configured to supply refrigerant liquid to the plurality of EUs and receive heated refrigerant vapor from the plurality of EUs, and further configured to dump heat from the heated refrigerant vapor into an HRU media that supplies heat to a reuse application, and an HRU controller including a non-transitory memory storage having programmed instructions, the programmed instructions implementing an iterative process for controlling the temperature of the processing unit packages. The process includes: 1) receiving temperature and frequency data from all processing units of a processing unit package; 2) determining that the number of processing units that are thermally and / or user throttled is less than a pre-defined maximum number of processing units during throttling (X1) and at least one processing unit is thermally throttled beyond a pre-defined thermal throttling upper limit (X2); 3) in response to the determination, sending a signal to the HRU condenser to increase the flow rate of the HRU media by a preset flow rate increment (X3); 4) thereafter, sending a signal to all processing units that are thermally throttled by at least the preset thermal throttling lower limit (X5) to reduce their user throttling maximum frequency by the preset user throttling increment (X4); 5) thereafter, sending a signal to the HRU condenser to reduce the flow rate of the HRU medium by a preset flow rate reduction increment (X6); 6) repeating the steps above until it is determined in step 2 that the number of thermally and user-throttled processing units is at least equal to a preset maximum number (X1) of processing units during throttling; may include:

[0008] The instructions may further include alternative processing of determining in step 2 that none of the processing units have been thermally throttled beyond a preset outlier thermal throttling limit (X2) and proceeding directly to step 5 in response to the determination.

[0009] Additionally, the instructions may further include maintaining the HRU media flow rate and user maximum processing unit frequency settings in response to determining in step 2 that the number of thermal and user throttled processing units is at least equal to a preset maximum number of throttling processing units (X1).

[0010] Additionally, the instructions may further include, in response to determining in step 2 that the number of thermally and user throttled processing units is at least equal to a preset maximum number of throttling processing units (X1), outputting processing unit temperature statistics including a mean and standard deviation of the processing unit temperatures.

[0011] In some embodiments, the preset maximum number of throttling processing units (X1) is preset at a level to maintain service level agreement (SLA) requirements. Further, in some embodiments, the preset maximum number of throttling processing units (X1) is set at 20-40 percent, such as 30 percent. The preset thermal throttling upper limit (X2) may be set at 5-20 percent, such as 10 percent, in some embodiments.

[0012] In some embodiments, the preset flow rate increase increment (X3) is set between 1 and 10 percent, for example 5 percent.

[0013] In some embodiments, the preset user throttling increment (X4) is set to 1-10 percent, for example 5 percent.

[0014] In some embodiments, the preset thermal throttling lower limit (X5) is set between 1 and 10 percent, such as 5 percent.

[0015] In some embodiments, the preset flow rate decrease increment (X6) is set to between 1 and 10 percent, such as 5 percent.

[0016] For a better understanding of various embodiments of the present invention and to show how they may be put into practice, reference will now be made by way of example to the accompanying drawings in which: Structural details of the invention are shown to provide a basic understanding of the invention, and the specification, taken together with the drawings, will make apparent to those skilled in the art how several forms of the invention may be embodied in practice. [Brief description of the drawings]

[0017] [Figure 1] 1 is a schematic diagram of a system for controlling heat reuse in a data center according to an embodiment of the present invention; [Figure 2A] FIG. 2 is a schematic diagram of heat flow in a system for heat reuse in a data center according to an embodiment of the present invention. [Figure 2B] FIG. 2 is a schematic diagram of heat flow in a system for heat reuse in a data center according to an embodiment of the present invention. [Diagram 3] FIG. 2 is a flow diagram of a process for controlling heat recycling in a data center according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] It is to be understood that the invention is not necessarily limited in its application to the details of construction and the arrangement of components and / or methods set forth in the following description and / or illustrated in the drawings, as the invention is capable of other embodiments or of being practiced or carried out in various ways.

[0019] An embodiment of the present invention provides a method for improving power efficiency of a data center. Exemplary embodiments of the present invention are described below. For clarity, not all features or components of an actual implementation are necessarily described. The embodiments and / or limitations shown in the drawings are selected for convenience or clarity of explanation and are not intended to limit the scope of the present invention.

[0020] In the following, the term "server" refers to a computer having one or more processing units (PUs), such as a central processing unit (processing unit) or a graphic processing unit (GPU). "Data center rack" is meant to denote a unit with which multiple servers are associated. A rack may be a frame in which multiple servers are mounted. However, in the following, a data center rack exemplifies any configuration of multiple servers in a data center drawing refrigerant from a single heat reuse (HRU) condenser, i.e., the configuration described below with respect to a rack may also apply to multiple racks, such as an "aisle" of a data center, or even to an entire data center, as described below.

[0021] Liquid cooling systems (e.g., water-based direct-on-chip systems) are generally configured to support a given "thermal design power" or "thermal design point" (TDP) that supports cooling of a processing unit operating at a maximum heat generation level. Even under "full load" conditions, the temperature and flow of the coolant (e.g., water or refrigerant) must be sufficient so that the processing unit temperature (i.e., the temperature of the junction between the processing unit chip and the processing unit package, referred to herein as "Tjunction") is maintained below a maximum value set by the manufacturer. Modern processing units are configured to implement "thermal throttling", i.e., to reduce energy consumption as their temperature approaches a maximum value. Typically, thermal throttling is implemented by reducing the operating clock speed, i.e., the operating frequency. Thermal throttling therefore reduces system performance.

[0022] Additionally, the servers described below are assumed to be configured to accept user instructions to modify the clock speed of the processing unit. Such user-initiated throttling, hereinafter referred to as user throttling, may modify any one or more of several clock parameters, such as cores, front side bus (FSB), and clock multipliers, and the voltage of the processing unit may also be modifiable to support the clock modifications and energy consumption levels.

[0023] Because the coolant system typically operates the same for both processing units operating at maximum levels and those that are only partially loaded, the coolant is typically heated to different temperatures by different processing units. Thus, when coolant flows from multiple processing units are mixed in a heat exchanger to release thermal energy, the resulting temperature is typically too low for effective heat reuse. Given the typical fluctuations in data center load, the coolant temperature also changes, making it impossible to predict the effective availability of heat for reuse.

[0024] It should be noted that different chip technologies and TDP values ​​result in different maximum temperatures (referred to herein as "Tcase") for the processing unit and processing unit package. The technology with the lowest maximum Tcase value is the technology for which the coolant temperature should be set so that the processing unit temperature does not exceed the maximum allowable value.

[0025] 1 is a schematic diagram of a system 100 for controlling heat reuse in a data center according to one embodiment of the present invention. The system 100 includes a data center 102 that supplies recycled heat to one or more external uses, such as a district heating plant 104. The district heating plant may be one or more buildings, such as a residential district.

[0026] The data center includes multiple server racks, such as server rack 106, on which multiple servers 114 are mounted. Each server includes one or more processing unit packages 116 that include a processing unit (also called a processing unit chip or "die"). When the processing units operate, they emit heat into the processing unit packages 116 that must be dissipated to prevent overheating. In an embodiment of the invention, a two-phase evaporator 118, also called evaporator or "evaporator unit" (EU) in the following, is attached to each processing unit package 116 to absorb the generated heat.

[0027] Refrigerant 120 circulates through the system 100, absorbing heat in the evaporator 118 and releasing heat in a heat reuse (HRU) condenser 130 (hereafter also referred to as a heat exchanger). As indicated by the arrows in the refrigerant cycle, when the refrigerant 120 enters the two-phase evaporator 118, the refrigerant is in a liquid state 120A. The heat absorbed in the evaporator 118 at least partially converts the refrigerant to a superheated state 120B (i.e., a partial or complete vapor).

[0028] From the evaporator 118, the refrigerant flows at a superheated state 120B to the HRU condenser 130 where it is returned to a liquid state 120A. Heat from the refrigerant is absorbed by an HRU medium 134, such as air or water, which provides the heat to the district heating plant 104. The HRU medium (also referred to herein as the condenser medium) enters the HRU condenser as a cold stream 134A and exits the district heating plant 104 as a heated stream 134B.

[0029] Furthermore, the system 100 includes an HRU controller 150, which may be any dedicated or general-purpose computer configured to control the operation of the HRU condenser 130 and the processing units of the processing unit package 116 according to data collected from the processing units. In particular, the HRU controller receives temperature and frequency data from all the processing units of the processing unit package 116. The HRU controller then manages the maximum processing unit frequency, as well as the flow rate of the HRU media 134, as described further below, to achieve an optimal heat generation level. Hereinafter, "optimal heat generation level" refers to a level of data center operation that generates waste heat at high temperature and efficient heat for reuse without causing an unacceptable degradation in the data center's performance. Typically, the operating parameters are at a level that results in some heat and user throttling, but does not prevent the data center from meeting service level agreement (SLA) requirements.

[0030] 2A and 2B are schematic diagrams of heat flow in the system 100 according to an embodiment of the present invention. Each diagram shows heat flow from a representative one of the processing unit (PU) packages 116 (i.e., one of the processing unit packages 116 described above). Heat generated by each processing unit package 116 is generated by a respective processing unit 202 enclosed within the processing unit package 116. The processing unit packages release heat to their respective evaporators 118. The evaporators then receive refrigerant from the HRU condenser 108. The evaporator 118 absorbs the heat of the processing unit package and turns the refrigerant into vapor. The vapor then flows to the HRU condenser 108, which then releases heat to a heat reuse application, such as heating the district heating plant 104, as described above.

[0031] 2A illustrates the heat flow of an exemplary scenario 200A in which the processing unit 202 is consuming 300 W of power, i.e., 300 W of heat must be dissipated through the processing unit package. Thus, the heat to be dissipated must be discharged through each subsequent stage of the heat flow until it is discharged to the district heating plant 104.

[0032] A given pressure of the refrigerant vapor is set by the HRU controller by controlling the condensation rate, which is done by controlling the flow of the HRU medium in the HRU condenser. The pressure of the vapor determines the boiling temperature in the evaporator. That is, the boiling temperature of the refrigerant can be adjusted by changing the pressure of the vapor in the refrigeration loop. Changing the flow rate of the HRU medium (which can be, for example, water or air) from the external cooling loop changes the condensation rate inside the heat exchanger. When the condensation rate in the heat exchanger is low, the pressure of the vapor increases and therefore the boiling temperature of the refrigerant in the evaporator becomes higher.

[0033] The temperature at the processing unit junction (referred to as "Tjunction") is shown in the exemplary scenario 200A to be approximately 90°C. With the thermal resistivity of the processing unit package estimated to be approximately 0.05°C / W, the temperature drop across the processing unit package is approximately (0.05°C / W) x (300W) = 15°C. That is, the temperature of the package, shown as "Tcase", is approximately 90°C - 15°C = approximately 75°C.

[0034] With the thermal resistivity of the evaporator estimated to be approximately 0.02°C / W, there is a temperature drop of approximately (0.02°C / W) x (300W) = 6°C. That is, as shown, the temperature of the refrigerant leaving the evaporator is approximately 69°C. The pressure of the refrigerant is typically adjusted so that 69°C is the evaporation temperature (i.e., boiling temperature), thereby providing a very efficient heat absorption at this temperature. As shown, the refrigerant at the inlet to the evaporator may be 62°C.

[0035] With a thermal resistivity of about 0.02° C. / W, there is a temperature drop in the HRU condenser of about (0.07° C. / W)×(300W)=about 2° C. This provides a sufficiently high temperature of about 2° C. for heat reuse.

[0036] As shown in FIG. 2B, in a second scenario 200B of a processing unit operating at part load and consuming only 60 W, the evaporator can operate in a similar refrigerant temperature range as above, i.e., with an input temperature of about 62° C. and an output temperature of about 69° C.

[0037] 3 is a flow diagram of a process 300 for controlling heat reuse in a data center according to one embodiment of the present invention. The steps of the process are typically performed by an HRU controller as described above. The process may be a continuous process or may be performed intermittently and stopped for a period of time once optimal operating conditions are achieved as described below. A system such as system 100 described above is envisioned in which an HRU condenser supplies refrigerant for multiple evaporator units (EU), each configured to operate as a two-phase evaporator. Typically, each EU is attached to a central processing unit or graphics processing unit (i.e., "PU") package and draws heat from the processing unit package. The condenser is configured to reject heat from the heated refrigerant to the HRU media, which provides heat for reuse applications.

[0038] As described below, a process implemented by the HRU controller gradually increases the temperature of all processing units regardless of their load while optimizing data center throughput performance. Without the process described below, some processing units may experience significant levels of thermal throttling while other processing units operate at lower temperature levels that prevent efficient heat reuse.

[0039] In a first step 310, the HRU controller receives temperature and frequency data from all processing units.

[0040] Based on the received data, in a subsequent step 320, the HRU controller determines whether a pre-set maximum number of processing units are thermally and / or user throttled. The pre-set maximum is typically set at a level that ensures that service level agreement (SLA) requirements can be met.

[0041] If the throttled number is less than the preset maximum number of throttled processing units (X1), processing continues to step 330. In this step, the HRU controller determines whether any of the throttled processing units are throttled beyond a preset thermal throttling upper limit (X2).

[0042] If any of the throttled processing units are throttled beyond the thermal throttling upper limit, processing proceeds to step 340 where the HRU controller signals the HRU condenser to increase the HRU media flow rate by a preset flow rate increment (X3) to reduce the refrigerant vapor pressure and thereby reduce the processing unit temperature, thereby reducing the extreme processing unit temperatures and allowing the processing unit temperatures to balance before further temperature increases are made.

[0043] Next, in step 350, the HRU determines which processing units are thermally throttled by the preset throttling lower limit (X5), and the HRU controller sends a signal to all such processing units to reduce their user throttling maximum frequency by the preset user throttling increment (X4).

[0044] Next, in step 360, the HRU sends a signal to the HRU condenser to decrease the flow rate of the HRU media by a preset flow rate decrease increment (X6). This step increases the pressure of the refrigerant vapor, thereby increasing the temperature of all processing units. By repeating this step multiple times for each cycle of the process (i.e., steps 310-360), the temperature of all processing units is gradually increased in a balanced manner to maximize the refrigerant temperature for heat reuse.

[0045] If in step 330, no processing units are throttled beyond a pre-set thermal throttling upper limit, step 340 may be skipped and processing may continue directly to step 350 (alternatively, processing may continue directly to step 360).

[0046] Steps 310-360 include step 320, which is a thermal and user throttling The process is repeated until it is determined that the number of processing units being throttled is at least equal to the preset maximum number of processing units during throttling. The HRU controller may then maintain the coolant and HRU media flow rates and maximum processing unit frequency in step 370 at least until a predetermined time has elapsed to repeat steps 310-360 to rebalance the processing unit temperatures. The HRU controller may also output statistics of the processing unit temperatures, such as the mean and standard deviation, as well as statistics regarding the processing unit frequency settings and HRU media flow rates to a log or to a system administrator.

[0047] For a typical data center, the preset maximum number of processing units during throttling (X1) may be set, for example, between 20 and 40 percent, with 30 percent being a typical value, and the preset thermal throttling upper limit (X2) may be set, for example, between 5 and 20 percent, with 10 percent being a typical value.

[0048] The preset flow rate increase increment (X3) may be set, for example, between 1 and 10 percent (5% is typical).

[0049] The preset user throttling increment (X4) may be set, for example, between 1 and 10 percent (typically, 5%).

[0050] The preset thermal throttling lower limit (X5) may be set, for example, between 1 and 10 percent (typically, 5%).

[0051] The preset flow rate reduction increment (X6) may be set, for example, between 1 and 10 percent (5% is typical).

[0052] As will be understood with reference to the above description and the referenced drawings, embodiments of computer-implemented methods are provided herein, some of which may be performed by various configurations of devices and systems described herein, some of which may be performed according to instructions stored on non-transitory computer-readable storage media. Furthermore, some embodiments of computer-implemented methods provided herein may be performed by other devices or systems and according to instructions stored on computer-readable storage media other than those described herein, as will be apparent to those skilled in the art with reference to the embodiments described herein. References to systems and computer-readable storage media in the following computer-implemented methods are provided for illustrative purposes, and none of such systems or any of such non-transitory computer-readable storage media are intended to be limiting with respect to the computer-implemented method embodiments described above. Similarly, any references to systems and computer-readable storage media in the following computer-implemented methods are provided for illustrative purposes, and are not intended to be limiting with respect to any of such computer-implemented methods disclosed herein.

[0053] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowcharts or block diagrams may represent a module, segment, or portion of code that includes one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks may occur in an order different from the order noted in the figures. For example, two blocks shown in succession may in fact be executed substantially simultaneously, or the blocks may be executed in reverse order, depending on the functionality involved. Furthermore, the blocks and / or flowcharts may be different from each other in the order of execution. It should also be noted that each block, and combinations of blocks in the block diagrams and / or flowchart illustrations, can be implemented by a special purpose hardware-based system that performs the specified functions or operations, or a combination of special purpose hardware and computer instructions.

[0054] The description of various embodiments of the present invention is presented for illustrative purposes, but is not intended to be exhaustive or limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terms used herein are selected to best explain the principles of the embodiments, practical applications, or technical improvements to the art found in the market, or to enable those skilled in the art to understand the embodiments disclosed herein. As used herein, the singular forms "a", "an", and "the" include plural references unless otherwise clear from the context.

[0055] As used herein, the term "exemplary" means "serving as an example, instance, or illustration," and should not necessarily be construed as preferred or advantageous over other embodiments or as excluding the incorporation of features from other embodiments.

[0056] It will be understood that certain features that are described for clarity in the context of separate embodiments may also be provided in combination in a single embodiment. Conversely, various features of the invention that are described for brevity in the context of a single embodiment may also be provided separately, in any suitable subcombination, or as appropriate with any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperable without those elements.

[0057] The above-described processes can be performed by software and / or hardware. These processes, or portions thereof, can be performed by computers, computer-type devices, workstations, processors, microprocessors, other electronic search tools and memories, and other non-transitory storage devices associated therewith. Furthermore, the processes, or portions thereof, can be embodied in programmable non-transitory storage media, such as compact discs (CDs) or other disks, including magnetic, optical, etc., readable by a machine or the like, other computer usable storage media, including magnetic, optical, or semiconductor storage devices, or other electronic signal sources.

[0058] The processes (methods) and systems, including their components, have been described herein with illustrative reference to specific hardware and software. The processes (methods) have been described by way of example, and those skilled in the art may omit and / or modify certain steps and their order in order to practice these embodiments without undue experimentation. The processes (methods) and systems have been described in a sufficient manner such that those skilled in the art may readily adjust other hardware and software, as necessary, to practice any embodiment, using conventional techniques without undue experimentation. The processing elements of the systems described herein may be implemented in digital electronic circuitry, or computer hardware, firmware, software, or combinations thereof. Such elements may be implemented as a computer program product tangibly embodied in an information carrier, such as a non-transitory machine-readable storage device, for execution by or to control the operation of a data processing device, such as a programmable processor or computer, or may be deployed to run on multiple computers at one location or on one or more computers across multiple locations. The memory storage device for the software and data may include a plurality of one or more memory units including one or more types of storage media. Examples of storage media include, but are not limited to, magnetic media, optical media, and integrated circuits such as read-only memory devices (ROM) and random access memories (RAM). A network interface module may control the transmission and reception of data packets over a network. The mobile device may be any computing device that allows user input to interactive applications as described above. Method steps associated with the systems and processes may be rearranged and / or one or more such steps may be omitted to achieve the same or similar results as those described herein.

[0059] It will be understood that the embodiments described herein above have been referred to by way of example, and that the present invention is not limited to what has been particularly shown and described herein above.

[0060] The following table defines the numerical designations used in the figures. [Table 1]

Claims

1. 1. A system for controlling heat reuse in a data center, comprising: a plurality of evaporator units (EUs), each configured to operate as a two-phase evaporator, each attached to a respective processing unit package of a data center server and adapted to draw heat from said respective processing unit package; a heat reuse unit (HRU) condenser configured to supply refrigerant liquid to the plurality of EUs and receive heated refrigerant vapor from the plurality of EUs, and further configured to reject heat from the heated refrigerant vapor to an HRU media that supplies heat to a reuse application; and an HRU controller including a non-transitory memory store having programmed instructions, the programmed instructions, when executed by the HRU controller, 1) receiving temperature and frequency data from all processing units of the processing unit package [step 310]; 2) The number of thermally and / or user-throttled processing units exceeds a preset maximum number of processing units during throttling (X １ ) and at least one processing unit is below a preset thermal throttling limit (X ２ determining that the system is thermally throttled beyond the threshold [steps 320 and 330]; 3) in response to said determining, causing said HRU condenser to increase the flow rate of the HRU media by a preset flow rate increment (X ３ ) [STEP 340]; and 4) Thereafter, the thermal throttling limit is at least the preset lower limit (X ５ ) to all processing units that are thermally throttled by the pre-defined user throttling increment (X ４ ) [STEP 350]; and 5) Thereafter, the flow rate of the HRU medium is decreased by a preset flow rate decrease increment (X ６ ) [STEP 360]; and 6) The number of processing units being thermally and user-throttled in step 2 is set to the preset maximum number of processing units during throttling (X １ repeating the steps until it is determined that the number of bits is at least equal to the number of bits in the first block (step 320); Run the system.

2. The instructions in step 2 cause any processing unit to exceed the preset outlier thermal throttling limit (X ２ 2. The system of claim 1, further comprising alternative processing of determining that the load has not been thermally throttled beyond step 10) and proceeding directly to step 5 in response to said determination.

3. The number of thermal and user throttled processing units is equal to or greater than the preset maximum value of throttling processing units (X １ 3. The system of claim 1, further comprising the step of: [Step 370] maintaining the HRU media flow rate and user maximum processing unit frequency settings in response to determining in Step 2 that the flow rate is at least equal to the user maximum processing unit frequency setting.

4. The number of thermal and user throttled processing units is equal to or greater than the preset maximum value of throttling processing units (X １ 2. The system of claim 1, further comprising the step of: in response to a determination in step 2 that the temperature of the process unit is at least equal to the mean and standard deviation of the process unit temperature, outputting statistics of the process unit temperature including the mean and standard deviation of the process unit temperature.

5. The preset maximum value of the processing units during throttling (X １ 2. The system of claim 1, wherein the bandwidth limit is preset to a level to maintain service level agreement (SLA) requirements.

6. The preset maximum value of the processing units during throttling (X １ ) is 30 percent 2. The system of claim 1, wherein the system is configured to

7. The preset thermal throttling upper limit (X ２ 2. The system of claim 1, wherein the threshold is set to 10 percent.

8. The preset flow rate increase increment (X ３ 2. The system of claim 1, wherein the threshold is set to 5 percent.

9. The preset user throttling increment (X ４ 2. The system of claim 1, wherein the threshold is set to 5 percent.

10. The preset thermal throttling lower limit (X ５ 2. The system of claim 1, wherein the threshold is set to 5 percent.

11. The preset flow rate reduction increment (X ６ 2. The system of claim 1, wherein the threshold is set to 5 percent.

12. The preset maximum value of the processing units during throttling (X １ 2. The system of claim 1, wherein the threshold is set to 20 to 40 percent.

13. The preset thermal throttling upper limit (X ２ 2. The system of claim 1, wherein the threshold is set to 5 to 20 percent.

14. The preset flow rate increase increment (X ３ 2. The system of claim 1, wherein the threshold is set to 1 to 10 percent.

15. The preset user throttling increment (X ４ 2. The system of claim 1, wherein the threshold is set to 1 to 10 percent.

16. The preset thermal throttling lower limit (X ５ 2. The system of claim 1, wherein the threshold is set to 1 to 10 percent.

17. The preset flow rate reduction increment (X ６ 2. The system of claim 1, wherein the threshold is set to 1 to 10 percent.

Images