WO2013129754A1

WO2013129754A1 - Method for virtualization-based control of cpu cooling, and computing apparatus for performing same

Info

Publication number: WO2013129754A1
Application number: PCT/KR2012/009080
Authority: WO
Inventors: 임성수; 오진수; 이호림
Original assignee: 국민대학교 산학협력단
Priority date: 2012-02-29
Filing date: 2012-10-31
Publication date: 2013-09-06
Also published as: KR101261354B1

Abstract

The method for the virtualization-based control of CPU cooling includes the steps of: (a) reducing the speed of a cooling fan corresponding to a particular CPU when said particular CPU is not consuming power due to a thermal effect; and (b) raising the speed of the cooling fan when the temperature of said particular CPU approaches a throttling temperature within a predetermined range, thereby giving priority to the prevention of CPU throttling rather than to the prevention of power consumption.

Description

Virtualization-based CPU cooling control method and computing device performing the same

The present application relates to a CPU cooling control technology, and more particularly, to a virtualization-based CPU cooling control method that can reduce power consumption through efficient CPU cooling and a computing device performing the same.

CPU (Central Processing Unit) cooling is performed to prevent performance degradation due to heat generation, and uses a CPU cooler. The CPU cooler is powered by a power source other than that provided to the CPU, and can be operated even when CPU cooling is not required.

Recently, computing devices have introduced virtualization to build at least one virtual machine on a single physical device. Here, a single physical device may include at least one CPU.

Korean Patent Laid-Open Publication No. 10-2001-0011151 uses a fast response thermoelectric module as a CAPI cooling element, so that CIF oil operates in a high output mode so that it can be quickly cooled even when the temperature rises rapidly. The method is disclosed.

Korean Patent Publication No. 10-2004-0052010 discloses an electronic chip cooling apparatus capable of maximizing a heat dissipation area by integrating a fan cover into a heat pipe by soldering or the like.

The present application is to provide a virtualization-based CPU cooling control method that can reduce power consumption through efficient CPU cooling and a computing device performing the same. For example, the present application may provide an appropriate trade off between increased performance by CPU cooling and power consumption by CPU cooler.

The present application is to provide a virtualization-based CPU cooling control method that can provide efficient CPU cooling through a power model in consideration of thermal effects based on regression analysis and a computing device performing the same.

Among the embodiments, the virtualization-based CPU cooling control method is a management virtual machine which is one of at least one virtual machine operated by a computing device including a CPU cooler implemented with at least one Central Processing Unit (CPU) and a cooling fan. Is performed in The method comprises the steps of (a) lowering the speed of a cooling fan corresponding to the particular CPU if power consumption does not occur due to thermal effects in the particular CPU; and (b) the temperature of the particular CPU is dependent on the CPU throttling temperature. Approaching a certain range includes prioritizing preventing the CPU throttling rather than preventing power consumption due to speeding up the cooling fan.

In an exemplary embodiment, the step (a) may further include lowering the speed of the cooling fan by a predetermined speed when the temperature of the specific CPU is lower than or equal to a first threshold temperature corresponding to a temperature at which power consumption due to the thermal effect occurs. It may include.

The step (b) may further include increasing the speed of the cooling fan by a predetermined speed if the temperature of the specific CPU is equal to or greater than a second threshold temperature associated with a CPU throttling temperature. The second threshold temperature may correspond to a predetermined temperature or less from the CPU throttling temperature.

In the step (b), if the temperature of the specific CPU is less than the second threshold temperature, the power consumption due to the increase of the speed of the cooling fan (hereinafter, referred to as the first power consumption) and the power consumption due to the thermal effect (hereinafter, referred to as 2 may be further included. The step (b) may further include increasing the speed of the cooling fan by the predetermined speed if the first power consumption is not higher than the second power consumption. The step (b) may further include maintaining a previous speed of the cooling fan if the first power consumption is higher than the second power consumption.

In one embodiment, the temperature at which power consumption occurs due to the thermal effect and the CPU throttling temperature may be predetermined for the particular CPU.

The method may further include repeating steps (a) and (b) after a predetermined time so as to confirm the cooling effect.

In embodiments, a computing device that performs virtualization-based CPU cooling control includes a CPU cooler implemented with at least one central processing unit (CPU) and a cooling fan, and includes a management virtual machine that is one of at least one virtualization machine. Virtualization-based CPU cooling control. The computing device may be configured to (a) reduce the speed of a cooling fan corresponding to the specific CPU if power consumption does not occur due to a thermal effect on the specific CPU, and (b) the temperature of the specific CPU may be reduced to a CPU throttling temperature. Approach gives priority to preventing the CPU throttling rather than preventing power consumption due to speeding up the cooling fan.

In embodiments, a computing device may include at least one central processing unit (CPU) that runs a management virtual machine, which is one of at least one virtual machine, and at least one CPU cooler corresponding to the at least one CPU and implemented as a cooling fan. Wherein the management virtual machine (a) lowers the speed of a cooling fan corresponding to the specific CPU if power consumption does not occur due to a thermal effect in the specific CPU, and (b) the temperature of the specific CPU is throttled by the CPU. Access to a throttling temperature performs virtualization-based CPU cooling control that prioritizes preventing CPU throttling rather than preventing power consumption due to speeding up the cooling fan.

The virtualization-based CPU cooling control method of the present application and the computing device performing the same can reduce power consumption through efficient CPU cooling.

The virtualization-based CPU cooling control method of the present application and a computing device performing the same may improve both power efficiency and performance of the computing device.

1 is a block diagram illustrating a virtualization-based computing device in accordance with one embodiment of the disclosed technology.

FIG. 2 is a block diagram illustrating the management virtual machine in FIG. 1.

3 is a flowchart illustrating a process of controlling CPU cooling based on virtualization by the management virtual machine of FIG. 1.

Description of the disclosed technology is only an embodiment for structural or functional description, the scope of the disclosed technology should not be construed as limited by the embodiments described in the text. That is, since the embodiments may be variously modified and may have various forms, the scope of the disclosed technology should be understood to include equivalents capable of realizing the technical idea. In addition, the objects or effects presented in the disclosed technology does not mean that a specific embodiment should include all or only such effects, and thus the scope of the disclosed technology should not be understood as being limited thereto.

On the other hand, the meaning of the terms described in the present application should be understood as follows.

Terms such as "first" and "second" are intended to distinguish one component from another component, and the scope of rights should not be limited by these terms. For example, the first component may be named a second component, and similarly, the second component may also be named a first component.

When a component is referred to as being "connected" to another component, it should be understood that there may be other components in between, although it may be directly connected to the other component. On the other hand, when a component is referred to as being "directly connected" to another component, it should be understood that there is no other component in between. On the other hand, other expressions describing the relationship between the components, such as "between" and "immediately between" or "neighboring to" and "directly neighboring", should be interpreted as well.

Singular expressions should be understood to include plural expressions unless the context clearly indicates otherwise, and terms such as "comprise" or "have" refer to features, numbers, steps, operations, components, parts, or parts thereof described. It is to be understood that the combination is intended to be present and does not exclude in advance the possibility of the presence or addition of one or more other features or numbers, steps, operations, components, parts or combinations thereof.

In each step, an identification code (e.g., a, b, c, etc.) is used for convenience of description, and the identification code does not describe the order of the steps, and each step clearly indicates a specific order in context. Unless stated otherwise, they may occur out of the order noted. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The disclosed technology can be embodied as computer readable code on a computer readable recording medium, and the computer readable recording medium includes all kinds of recording devices in which data can be read by a computer system. . Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like, and are also implemented in the form of a carrier wave (for example, transmission over the Internet). It also includes. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. Generally, the terms defined in the dictionary used are to be interpreted to coincide with the meanings in the context of the related art, and should not be interpreted as having ideal or excessively formal meanings unless clearly defined in the present application.

Referring to FIG. 1, a virtualization-based computing device 100 includes a CPU pool 110, a memory 120, and at least one including at least one central processing unit (CPU) 112 and a CPU cooler 114. Virtual machine 130.

The CPU pool 110 includes at least one CPU 112 and a CPU cooler 114 and corresponds to a resource allocated or returned by an operating system or the like. For example, if a particular task requires a large amount of computation, the operating system can allocate a relatively large number of CPUs to a particular task. If a particular task requires a small amount of computation, the operating system allocates a relatively small number of CPUs for a particular task. can do. Meanwhile, the CPU cooler 114 may be implemented as a cooling fan, and the speed of the cooling fan may be controlled according to the heat generated by the CPU 112. Control of such a cooling fan will be described with reference to FIGS. 2 and 3.

Memory 120 may be used as shared memory for communication between at least one CPU 112 or communication between at least one CPU 112 and another device, or as local memory for data storage by at least one CPU 112. have. In one embodiment, the memory 120 may be implemented as a volatile memory, a nonvolatile memory, or a combination thereof.

The at least one virtual machine 130 may be implemented in software operated by the computing device 100. Each of the at least one virtual machine 140 may be recognized as a separate physical device from the user's point of view, and one of the at least one virtual machine 140 corresponds to the management virtual machine 132 that controls CPU cooling. . That is, the management virtual machine 132 corresponds to one of the at least one virtual machine 140 and controls the speed of the cooling fan in the CPU cooler 120.

2, the management virtual machine 132 includes a CPU monitoring module 210, a power model management module 220, a CPU cooler monitoring module 230, and a management module 240.

The CPU monitoring module 210 monitors the temperature occurring in the specific CPU 112a. In one embodiment, the temperature may be measured through a sensor built in the specific CPU 112a or an externally mounted sensor, and the CPU monitoring module 210 may measure the temperature generated by the specific CPU 112a through the sensor. I can recognize it.

The control module 220 presents an appropriate power model between increasing the performance of a particular CPU 112a by CPU cooling and power consumption by the CPU cooler 114. More specifically, the control module 220 proposes a power model capable of performing efficient CPU cooling through a power model in consideration of thermal effects based on regression analysis. To this end, the control module 220 may manage three data structures. The first data structure is used to manage information relating to a first threshold temperature of each of the at least one CPU 112, and the second data structure provides information about power consumption according to a current temperature of each of the at least one CPU 112. The third data structure is used to manage information on power consumption according to the speed of the cooling fan in the at least one CPU cooler 114. All of the first to third data structures may correspond to static tables, and the first threshold temperature, power consumption according to the current temperature, and power thahfifd according to the cooling fan may be predetermined for the specific CPU 112a. .

The cooler monitoring module 230 monitors the speed of the particular CPU cooler 114a. In one embodiment, the cooler monitoring module 230 may monitor the speed of the particular CPU cooler 114a based on the amount of power consumed by the cooling fan.

The management module 240 provides an interface to the management virtual machine 132 to monitor or control the CPU 112 and the CPU cooler 114. That is, the management module 240 may serve as an interface for determining the temperature of the specific CPU 112a or the speed of the specific CPU cooler 114a, and serve as an interface for controlling the speed of the specific CPU cooler 114a. Can be performed.

Hereinafter, a process of controlling CPU cooling based on virtualization by the management virtual machine 132 will be described with reference to FIG. 3.

In FIG. 3, the control module 220 waits for a specific time (eg, 5 minutes) to confirm the CPU cooling effect (step S310). In one embodiment, the specific time may be fixed. This fixing may allow the cooling control process to be performed periodically. In another embodiment, the specific time may vary. Such a variable may be determined based on a temperature change range of the specific CPU 112a. For example, if the temperature change range is above a certain threshold (eg, 20 degrees), the control module 220 may immediately stop waiting.

The control module 220 may obtain the current temperature of the specific CPU 112a from the CPU monitoring module 210 (step S315). In one embodiment, the CPU monitoring module 210 may continue to measure the temperature for the at least one CPU 112 through the management module 240, and in response to the request of the control module 220, the specific CPU 112a. Can provide the current temperature for. Here, the CPU monitoring module 210 may maintain a current temperature table for each of the at least one CPU 112. In another embodiment, the CPU monitoring module 210 may provide a current temperature of the specific CPU 112a after measuring the current temperature of the specific CPU 112a according to a request of the control module 220.

The control module 220 determines whether the current temperature of the specific CPU 112a exceeds a temperature at which power consumption occurs due to a thermal effect in the specific CPU 112a (hereinafter, referred to as a first threshold temperature). (Step S320). That is, the control module 220 determines whether power consumption due to a thermal effect occurs in the specific CPU 112a. In one embodiment, the control module 220 may manage a first threshold temperature for each of the at least one CPU 112 and, if a current temperature of the particular CPU 112a is received, the current temperature of the particular CPU 112a. And the first threshold temperature of the specific CPU 112a may be compared.

On the other hand, as an embodiment different from steps S315 and S316, the CPU monitoring module 210 may manage a first threshold temperature for each of the at least one CPU 112, and the specific CPU 112a through the management module 240. ) Can be obtained to compare the current temperature of the specific CPU 112a with the first threshold temperature of the specific CPU 112a. Next, the CPU monitoring module 210 may notify the control module 220 of such excess information if the current temperature of the specific CPU 112a exceeds the first threshold temperature of the specific CPU 112a.

The control module 220 supplies the CPU cooler 114a corresponding to the specific CPU 112a through the management module 240 if the current temperature of the specific CPU 112a does not exceed the first threshold temperature of the specific CPU 112a. The speed of the cooling fan is lowered (step S325). In one embodiment, the control module 220 may lower the speed of the cooling fan by a predetermined speed (eg, 100 RPM). The cooler monitoring module 230 monitors whether the speed of the cooling fan is lowered through the management module 240 to determine whether to change the speed of the cooling fan (step S380). In one embodiment, the cooler monitoring module 230 may determine whether to change the speed of the cooling fan based on the amount of power consumed by the cooling fan.

As a result, in steps S320 and S325 described above, the control module 220 lowers the speed of the cooling fan corresponding to the specific CPU 112a if power consumption due to a thermal effect does not occur in the specific CPU 112a.

If the current temperature of the specific CPU 112a exceeds the first threshold temperature of the specific CPU 112a, the control module 220 may have a current temperature within a predetermined range below the CPU throttling temperature. A second threshold temperature) is determined (step S330). For example, the predetermined range may correspond to 5 degrees, and the predetermined range may be set to prevent CPU throttling from occurring. Here, CPU throttling is known as dynamic voltage scaling, and refers to a technology in which the frequency of the CPU is automatically adjusted to conserve power or reduce the amount of heat generated by the CPU.

The control module 220 speeds up the cooling fan in the CPU cooler 114a corresponding to the specific CPU 112a through the management module 240 when the current temperature of the specific CPU 112a is equal to or greater than the second threshold temperature (see FIG. Step S335). In one embodiment, the control module 220 may increase the speed of the cooling fan by a predetermined speed (eg, 100 RPM). The cooler monitoring module 230 monitors whether the speed of the cooling fan is increased through the management module 240 to determine whether to change the speed of the cooling fan (step S380).

The control module 220 controls the power consumption of the specific CPU 112a and the power consumption of the specific CPU cooler 114a due to the thermal effect when the current temperature of the specific CPU 112a is less than the second threshold temperature (that is, the speed of the cooling fan). The power consumption due to the increase) is measured (step S340). The power consumption of the particular CPU 112a may be determined based on the current temperature measured by the CPU monitoring module 210. The power consumption of the particular CPU cooler 114a may be measured by the cooler monitoring module 230.

The control module 220 determines whether the power consumption of the specific CPU cooler 114a is higher than the power consumption of the specific CPU 112a (step S350). If so, the control module 220 maintains the speed of the cooling fan in the specific CPU cooler 114a (step S360). If not, the control module 220 speeds up the cooling fan in the specific CPU cooler 114a (step S355).

As a result, in steps S330, S340, S350, and S360, the control module 220 determines that the cooling fan of the cooling fan in the specific cooler 114a if the temperature of the specific CPU 112a approaches the CPU throttling temperature within a certain range. Priority is given to preventing CPU throttling rather than preventing power consumption from speeding up.

On the other hand, if the speed change of the cooling fan fails in step S380, the control module 220 determines whether the speed of the cooling fan is the maximum or minimum (step S365). If so, the control module 220 maintains the speed of the cooling fan in the specific CPU cooler 114a (step S360), otherwise, the control module 220 controls the speed of the cooling fan again (step S370). It is determined whether or not the fan speed change is successful (step S380).

On the other hand, the steps S310 to S380 are repeated after a predetermined time elapsed in step S310 so as to confirm the cooling effect.

Although described above with reference to a preferred embodiment of the present application, those skilled in the art various modifications and changes to the present application without departing from the spirit and scope of the present application described in the claims below I can understand that you can.

Claims

A virtualization-based CPU cooling control method performed in a management virtual machine which is one of at least one virtual machine operated by a computing device including a CPU cooler implemented with at least one central processing unit (CPU) and a cooling fan,

(a) lowering a speed of a cooling fan corresponding to the specific CPU if power consumption due to a thermal effect does not occur in the specific CPU; And

(b) if the temperature of the particular CPU approaches the CPU throttling temperature within a certain range, the priority is given to preventing the CPU throttling rather than preventing power consumption due to speeding up the cooling fan. Virtualization-based CPU cooling control method comprising the steps.
The method of claim 1, wherein step (a)

Virtualization-based CPU cooling, further comprising: lowering the speed of the cooling fan by a predetermined speed when the temperature of the specific CPU is lower than a first threshold temperature corresponding to a temperature at which power consumption due to the thermal effect occurs. Control method.
The method of claim 2, wherein step (b)

And if the temperature of the particular CPU is greater than or equal to a second threshold temperature associated with a CPU throttling temperature, increasing the speed of the cooling fan by a predetermined speed.
4. The method of claim 3, wherein the second threshold temperature is

Virtualization-based CPU cooling control method characterized in that it corresponds to a predetermined temperature or less from the CPU throttling temperature.
The method of claim 3, wherein step (b)

When the temperature of the specific CPU is less than the second threshold temperature, the power consumption due to the speed of the cooling fan (hereinafter, referred to as first power consumption) and the power consumption due to the thermal effect (hereinafter referred to as second power consumption) are measured. Virtualization-based CPU cooling control method further comprising the step of.
The method of claim 5, wherein step (b)

And increasing the speed of the cooling fan by the predetermined speed if the first power consumption is not higher than the second power consumption.
The method of claim 6, wherein step (b)

And maintaining the transfer speed of the cooling fan if the first power consumption is higher than the second power consumption.
The method of claim 1, wherein the temperature at which power consumption due to the thermal effect occurs and the CPU throttling temperature is

Virtualization-based CPU cooling control method characterized in that the predetermined for the particular CPU.
The method of claim 1,

Virtualization-based CPU cooling control method further comprises the step of repeating the step (a) and (b) after a predetermined time so that the cooling effect can be confirmed.
A computing device including a CPU cooler implemented with at least one central processing unit (CPU) and a cooling fan, and operating a management virtual machine, which is one of at least one virtualization machine, to perform virtualization-based CPU cooling control. Computing devices

(a) if power consumption does not occur due to thermal effects in a particular CPU, lower the speed of the cooling fan corresponding to the specific CPU;

(b) if the temperature of the particular CPU approaches a CPU throttling temperature, prioritizing preventing the CPU throttling rather than preventing power consumption due to speeding up the cooling fan.
At least one central processing unit (CPU) to operate a management virtual machine, which is one of the at least one virtualization machine; And

At least one CPU cooler corresponding to the at least one CPU and implemented as a cooling fan,

The management virtual machine (a) lowers the speed of the cooling fan corresponding to the specific CPU if the power consumption does not occur due to thermal effects in the specific CPU, and (b) the temperature of the specific CPU is the CPU throttling temperature And a virtualization-based CPU cooling control that gives priority to preventing the CPU throttling rather than preventing power consumption due to speeding up the cooling fan.