WO2017016359A1 - 运算资源的控制方法、装置和计算机存储介质 - Google Patents

运算资源的控制方法、装置和计算机存储介质 Download PDF

Info

Publication number
WO2017016359A1
WO2017016359A1 PCT/CN2016/087227 CN2016087227W WO2017016359A1 WO 2017016359 A1 WO2017016359 A1 WO 2017016359A1 CN 2016087227 W CN2016087227 W CN 2016087227W WO 2017016359 A1 WO2017016359 A1 WO 2017016359A1
Authority
WO
WIPO (PCT)
Prior art keywords
computing resources
computing
computing resource
resource
temperature
Prior art date
Application number
PCT/CN2016/087227
Other languages
English (en)
French (fr)
Inventor
张文彦
Original Assignee
深圳市万普拉斯科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市万普拉斯科技有限公司 filed Critical 深圳市万普拉斯科技有限公司
Priority to US15/747,482 priority Critical patent/US10621011B2/en
Priority to EP16829725.7A priority patent/EP3330834B1/en
Publication of WO2017016359A1 publication Critical patent/WO2017016359A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5094—Allocation of resources, e.g. of the central processing unit [CPU] where the allocation takes into account power or heat criteria
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/16—Constructional details or arrangements
    • G06F1/20—Cooling means
    • G06F1/206—Cooling means comprising thermal management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3206—Monitoring of events, devices or parameters that trigger a change in power modality
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
    • G06F9/4893—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues taking into account power or heat criteria
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Definitions

  • the present invention relates to the field of electronic technologies, and in particular, to a method, an apparatus, and a computer storage medium for controlling computing resources.
  • Embodiments of the present invention are directed to a method, apparatus, and computer storage medium for computing resource control to solve the above problems.
  • An embodiment of the present invention provides a method for controlling a computing resource, where the method includes: acquiring a highest frequency that the first computing resource in the device supports to run; and determining, according to a predetermined rule, a number of cold rooms corresponding to the highest frequency;
  • the determining the number of cold rooms corresponding to the highest frequency according to the predetermined rule comprises: searching for a preset frequency point matching the highest frequency according to the preset cold room demand table, and obtaining the cold room demand table The number of the cold rooms corresponding to the frequency points, wherein the correspondence between the preset frequency points and the number of cold rooms is recorded in the cold room demand table.
  • the determining the specific number of second computing resources, the step of turning off the specific number of second computing resources comprises: determining idle computing resources from the plurality of second computing resources, high load and long time Use the computing resources and the least busy computing resources; turn off the computing resources according to one of the following: select and close a specific number of idle computing resources; turn off all idle computing resources, and partially high-load and long-term computing resources, The sum of the number of idle computing resources and the high-load and long-term computing resources is equal to the specific number; all idle computing resources and all high-load and long-term computing resources, and some of the least busy computing resources are closed. The sum of the idle computing resources, the high load and long used computing resources, and the least busy computing resources is equal to the specific number.
  • the method further includes: periodically acquiring the temperature of each computing resource; and detecting that the computing resource has a high temperature When the preset value is used, the computing resource whose temperature is higher than the preset value is turned off, and the task to be processed by the closed computing resource is processed by the computing resource with the lowest temperature among the computing resources that are closed according to the number of cold rooms.
  • the method further includes: determining a maximum frequency that each computing resource supports operation, from the determined Determining a minimum value in the highest frequency, determining a number of cold rooms corresponding to the minimum value according to a predetermined rule, determining whether the determined number of cold rooms is less than a quantity of computing resources that have been currently closed; if the determined number of cold rooms is less than The number of closed computing resources further determines whether the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost; if so, enabling a closed computing resource.
  • An embodiment of the present invention further provides a method for controlling a computing resource, where the method includes: periodically acquiring temperatures of all computing resources;
  • the computing resource whose temperature is higher than the preset value is turned off, and the processing required for the closed computing resource is transferred to the computing resource with the lowest temperature among the closed computing resources. Process it.
  • the embodiment of the present invention further provides an apparatus for computing resource control, the apparatus comprising: an obtaining module configured to acquire a highest frequency supported by a first computing resource in the device; and a determining module configured to determine according to a predetermined rule a number of cold rooms corresponding to the highest frequency; a shutdown module configured to determine a specific number of second computing resources to shut down the specific number of second computing resources, wherein the specific number is the same as the number of the cold rooms.
  • the determining module is further configured to: according to the preset cold room demand table, find a preset frequency point that matches the highest frequency, and obtain a number of cold rooms corresponding to the frequency point in the cold room demand table, The correspondence between the preset frequency point and the number of cold rooms is recorded in the cold room demand table.
  • the shutdown module includes: a type determining module configured to determine an idle computing resource, a high-load and long-term computing resource, and a least busy computing resource from the plurality of second computing resources; Close the module and configure it to shut down the computing resources as follows:
  • the apparatus further includes: a temperature acquisition module configured to periodically acquire temperature of all computing resources; and a switching module, when detecting that the temperature of the computing resource is higher than a preset value, the shutdown temperature is higher than the preset The value of the computing resource, the task to be processed by the closed computing resource is processed by the lowest temperature computing resource in the closed computing resource.
  • the apparatus further includes: a first determining module, configured to determine a highest frequency that each computing resource supports to run, determine a minimum value from the determined highest frequency, and determine and describe according to a predetermined rule The number of the cold rooms corresponding to the minimum value, determining whether the determined number of cold rooms is less than the number of computing resources that have been closed; the second determining module, if the determined number of cold rooms is less than the number of computing resources that have been closed, Further determining whether the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost; the resource enabling module is configured to spend less than the current maximum current cost if a closed computing resource is enabled, Then enable a closed computing resource.
  • a first determining module configured to determine a highest frequency that each computing resource supports to run, determine a minimum value from the determined highest frequency, and determine and describe according to a predetermined rule The number of the cold rooms corresponding to the minimum value, determining whether the determined number of cold rooms is less than the number of computing resources that have been closed; the second
  • the embodiment of the invention further provides a control device for computing resources, the device comprising:
  • a temperature acquisition module configured to periodically acquire temperature of all computing resources
  • the switching module is configured to: when detecting that the temperature of the computing resource is higher than a preset value, turning off the computing resource whose temperature is higher than the preset value, and transferring the task to be processed by the closed computing resource to the temperature of the closed computing resource The lowest computing resources are processed.
  • the embodiment of the invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the control method of the computing resource according to the embodiment of the invention.
  • the method and device for controlling the computing resources and the computer storage medium determine the number of cold rooms corresponding to the highest frequency according to a predetermined rule by acquiring the highest frequency that the first computing resource is currently operable in the device. Determining a certain number of second computing resources, the specific number of second computing resources being closed, wherein the specific number is the same as the number of the cold rooms. Based on the characteristics of heat that is spontaneously transmitted from a high temperature object to a low temperature object, the first When the temperature of the second computing resource is reduced, the heat of the first computing resource with a higher temperature is transferred to the second computing resource with a lower temperature, thereby achieving heat dissipation of the first computing resource.
  • 1 is a flow chart of a method for computing resource control in an embodiment
  • FIG. 3 is a flow diagram of enabling a closed computing resource in one embodiment
  • FIG. 5 is a structural block diagram of an apparatus for computing resource control in an embodiment
  • FIG. 6 is a structural block diagram of a shutdown module in an embodiment
  • FIG. 7 is a structural block diagram of an apparatus for computing resource control in another embodiment
  • FIG. 8 is a structural block diagram of another apparatus for computing resource control in an embodiment.
  • a method for computing resource control is proposed, which can be applied to various electronic devices, including but not limited to personal computers, notebook computers, smart phones, tablets, The smart wearable device or the like, the method specifically includes:
  • Step 110 Acquire a highest frequency that the first computing resource in the device supports running.
  • the computing resource may be a core of a central processing unit (CPU) in the processor.
  • the existing typical processing unit for the mobile terminal may include four, six or eight central processing units. core.
  • the computing resource may also be a Graphics Processing Unit (GPU) in the processor. It will be appreciated that those skilled in the art will appreciate that the computing resources are not limited to the foregoing two.
  • the The computing resources can be computationally capable components such as cores, main cores, sub-cores, hardware engines, etc. in the processor.
  • the above computing resources may be a single one of the above, or a combination of the above.
  • the computing resources in the device generally support multi-frequency point operations, and generally select an appropriate frequency point as the operating frequency among all the selectable frequency points according to the calculated load amount and usage rate.
  • the highest frequency is the one that is the largest of all the optional frequency points.
  • the thermal manager will force a lower frequency point at which the computing resources can be selected when high temperatures are detected, and the highest frequency will also decrease.
  • frequency points available 634MHz, 768MHz, 864MHz, 960MHZ, 1248MHz, 1344MHz, 1440MHz, 1536MHz, 1632MHz, 1766MHz.
  • the highest frequency is 1766MHz, when the computing resource temperature is high.
  • the available frequency points are gradually decreased. For example, the frequency is reduced to only 9 frequency points.
  • the highest frequency is 1632MHz. If the temperature continues to rise, the frequency point drops to only 8 available. The highest frequency is reduced. 1536MHz, and so on.
  • Step 120 Determine, according to a predetermined rule, the number of cold rooms corresponding to the highest frequency.
  • the cold room demand table is preset and stored in the device, and the correspondence relationship between the preset frequency point and the number of cold rooms is recorded in the cold room demand table. Specifically, the preset frequency point matching the highest frequency in the cold room demand table is searched, and the number of cold rooms is determined according to the preset frequency point.
  • the frequency points recorded in the cold room demand table are frequency critical points corresponding to the number of different cold rooms. When the highest frequency is between the two frequency critical points, the frequency point matching the highest frequency is larger. That frequency point.
  • the number of cold rooms can be calculated by a particular formula based on the maximum frequency of computing resources.
  • a processor has four computing resources available, providing frequencies of 384 MHz, 480 MHz, 768 MHz, 864 MHz, 960 MHz, 1248 MHz, 1344 MHz, 1632 MHz, and 1766 MHz.
  • the preset cold room demand table can be as shown in Table 1.
  • the frequency point matching the current highest frequency is 1248MHz, and the corresponding number of cold rooms is 1.
  • the frequency point matching the current highest frequency is 1766MHz.
  • the frequency point matching the current highest frequency is 768MHz.
  • Step 130 Determine a specific number of second computing resources, and shut down the specific number of second computing resources, wherein the specific quantity is the same as the number of cold rooms.
  • This embodiment performs heat dissipation by turning off the second computing resource. Specifically, the second computing resource is turned off, that is, the second computing resource is powered off, and the temperature of the second computing resource is decreased. When the temperature of the second computing resource is lower than the first computing resource, considering the directivity of heat conduction, that is, the heat is spontaneously transmitted from the high temperature object to the low temperature object, and the heat of the first computing resource having a higher temperature is lower to the temperature. The second computing resource is delivered to implement heat dissipation of the first computing resource. The second computing resource that is turned off is therefore referred to as a cold room.
  • the second computing resource and the first computing resource have the same type, and are all cores of the central processing unit in the processor. It can be understood that, in other implementations, the categories of the second computing resource and the first computing resource may be different when needed, for example, the first computing resource is a core of a central processing unit in the processor, and the second computing resource is It can be one or a combination of components of computing power, such as a graphics processing unit, a hardware engine, or the like in a processor.
  • Selecting a specific number of second computing resources from the plurality of second computing resources is not random, and one of a plurality of factors to be considered is a physical layout of the second computing resource, that is, a second from the first computing resource. Select from computing resources.
  • the step of determining a specific number of second computing resources and shutting down a specific number of second computing resources includes:
  • Step 130a Determine idle computing resources, high-load and long-term computing resources, and least busy computing resources from the plurality of second computing resources.
  • the types of computing resources are classified into three types in advance, namely, idle computing resources, high-load and long-term computing resources, and least busy computing resources.
  • the idle computing resource refers to an computing resource that has no computing task and is in an idle state.
  • the high-load and long-term computing resource refers to a computing resource that has been busy for a period of time, and the least busy computing resource refers to a running load. The lowest computing resource.
  • step 130b the computing resource is turned off according to one of the following:
  • the computing resources equal to the number of cold rooms are turned off according to the number of cold rooms. Specifically, if the quantity satisfies the requirement, the idle computing resource is preferentially closed. When the number of idle computing resources is less than the number of cold rooms, all idle computing resources are turned off, and then some high-load and closed computing resources are used for a long time. When all the idle computing resources and the total number of high-load and long-term computing resources are less than the number of cold rooms, both of them are closed, and then some of the least busy computing resources are closed.
  • the remaining three choose to turn off the high-load and long-term computing resources. If the number of computing resources that are heavily loaded and used for a long time is less than three, the remaining reselection turns off the least busy computing resources.
  • the idle computing resource is preferentially turned off because it has no computing task, but the clock source and power supply configured for it are not completely disconnected, but only enter the lower current mode, and the power that is turned off is completely broken. Open, so there is no possibility of any electricity being converted into heat, and the highest priority is to help the heat dissipation, so it is most appropriate to turn off the idle computing resources first. If the number of idle computing resources is insufficient, then the computing resources that are used to turn off the high load and use for a long time are selected because the high load continues for a period of time, and the temperature is higher.
  • the highly loaded and long-running computing resources are turned off as the second priority object. If the number of the above two still does not meet the requirements, choose to turn off the least busy computing resources, the temperature is in a neutral state, there is no need to close the need, so stay at the end. By properly selecting the order in which the computing resources are turned off, heat can be dissipated more efficiently.
  • a cold room rotation can also be performed. After the step 130 of determining the specific number of second computing resources to turn off the specific number of second computing resources, the method further includes:
  • the temperature of each computing resource is obtained periodically.
  • a temperature sensor is disposed in the vicinity of each computing resource, and the temperature of each computing resource can be quickly detected.
  • the data of the temperature rise sensor can be obtained periodically (for example, 1 second) to determine the temperature of each computing resource.
  • the shutdown temperature is higher than a preset value.
  • the computing resource, the task to be processed by the closed computing resource is processed by the lowest temperature computing resource in the closed computing resource.
  • the computing resources in the above steps all belong to the same category, for example, they are cores in the central processing unit of the processor.
  • a preset value is preset for the temperature of each computing resource, and after the computing resource is closed according to the number of cold rooms, if the temperature of the computing resource is found to be relatively high, but it is not enough to close another computing resource, Then, when it is detected that the temperature of a computing resource is higher than a preset value, the computing resource is closed, and the task required to be operated by the closed computing resource is assigned to the lowest temperature operation of the currently closed computing resource. Resources are processed. In this way, by turning off the high-temperature computing resources and simultaneously enabling the closed-lowest computing resources, the heat dissipation can be more uniform and uniform.
  • the closed computing resources can be re-enabled to be restored into computing resources that can be coordinated.
  • the method further includes:
  • Step 142 determining a maximum runable frequency of each computing resource, determining a minimum value from the determined highest frequency, determining a number of cold rooms corresponding to the minimum value according to a predetermined rule, and determining whether the determined number of cold rooms is It is less than the number of computing resources that have been currently closed, and if so, proceeds to step 144, otherwise ends.
  • the currently running computing resources share the same clock source, the highest running frequency of the currently running computing resource is the same, and the minimum value is currently running. The highest frequency at which a computing resource can run.
  • the highest operable frequency of each computing resource is determined again, and a minimum value is determined from the highest frequencies of all computing resources, according to a preset cold room demand table. Find the preset frequency in the cold room demand table that matches the minimum value Rate point, and then according to the preset frequency point, find the corresponding number of cold rooms in the cold room demand table.
  • the frequency points recorded in the cold room demand table are frequency critical points corresponding to the number of different cold rooms. When the minimum value is between the two frequency critical points, the frequency point matching the minimum value is the smaller frequency point.
  • the current number of cold rooms is 2, and when the minimum value returns to above 1248 MHz, a cold room is allowed to be released (ie, a closed computing resource is enabled), and 1248 MHz is the minimum frequency point that must be reached to release a cold room at this time.
  • step 144 it is determined whether the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost, and if so, proceeds to step 306, otherwise ends.
  • the current consumption is power consumption
  • the calculation formula is Power(f)*M, where f represents frequency, Power(f) represents power consumption at frequency f, and M represents the number of computing resources being operated.
  • the minimum current cost refers to the current cost of the frequency at the minimum frequency.
  • the minimum current cost of enabling a closed computing resource can be expressed as Power(f min )*(M+1), where f min represents the use of a closed
  • the current maximum current cost is expressed as Power(F)*M, which represents the current cost of the current highest frequency F in the case where the number of cold rooms M is constant, where F represents the highest frequency currently allowed. Determining whether the minimum current cost after using a closed computing resource is lower than the current maximum current cost, that is, determining whether the current maximum power consumption is sufficient to withstand the power consumption after releasing a cold room, and if so, proceeding Step 146, if no, then ends.
  • Step 146 enabling a closed computing resource.
  • the current number of cold rooms is greater than the required number of cold rooms, and the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost, then releasing a cold room to enable one is turned off.
  • Computing resources In this way, the computing resources can be fully and reasonably utilized without causing waste of computing resources.
  • the embodiment of the invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the control method of the computing resource according to the embodiment of the invention.
  • a method for controlling a computing resource comprising:
  • step 202 the temperature of all computing resources is obtained periodically.
  • a temperature sensor is disposed in the vicinity of each computing resource, and the temperature of each computing resource can be quickly detected.
  • the data of the temperature rise sensor can be obtained periodically (for example, 1 second) to determine the temperature of each computing resource.
  • Step 204 When it is detected that the temperature of the specific computing resource is higher than the preset value, the computing resource whose temperature is higher than the preset value is turned off, and the task to be processed by the closed computing resource is left to the lowest temperature of the closed computing resource.
  • the computing resources are processed.
  • the computing resources in the above steps 202 and 204 all belong to the same category, for example, they are cores in the central processing unit of the processor.
  • a preset value is preset for the temperature of each computing resource, and after the computing resource is closed according to the number of cold rooms, if the temperature of the computing resource is found to be relatively high, but it is not enough to close another computing resource, Then, when it is detected that the temperature of a computing resource is higher than a preset value, the computing resource is closed, and the task required to be operated by the closed computing resource is assigned to the lowest temperature operation of the currently closed computing resource. Resources are processed. In this way, by turning off the high-temperature computing resources and simultaneously enabling the closed-lowest computing resources, the heat dissipation can be more uniform and uniform.
  • steps 202 and 204 above are set after step 130.
  • the embodiment of the invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the control method of the computing resource according to the embodiment of the invention.
  • a control device for computing resources comprising:
  • the obtaining module 410 is configured to acquire a highest frequency that the first computing resource in the device supports running.
  • the determining module 420 is configured to determine the number of cold rooms corresponding to the highest frequency according to a predetermined rule.
  • the closing module 430 is configured to determine a specific number of second computing resources to shut down a specific number of second computing resources, wherein the specific number is the same as the number of cold rooms.
  • the determining module 420 is further configured to: according to the preset cold room demand table, find a preset frequency point that matches the highest frequency, and obtain a number of cold rooms corresponding to the frequency point in the cold room demand table, The correspondence between the preset frequency point and the number of cold rooms is recorded in the cold room demand table.
  • the shutdown module 430 includes:
  • the type determination module 430a is configured to determine an idle computing resource, a high load and long time used computing resource, and a least busy computing resource from the plurality of second computing resources.
  • the resource shutdown module 430b is configured to shut down the computing resource according to one of the following:
  • the above apparatus further includes:
  • the temperature acquisition module is configured to periodically acquire the temperature of each computing resource.
  • the switching module is configured to: when detecting that the temperature of a computing resource is higher than a preset value, shutting down the computing resource whose temperature is higher than a preset value, and handing off the task to be processed by the closed computing resource to the closed The lowest temperature computing resource in the computing resource is processed.
  • the foregoing apparatus further includes:
  • the first determining module 440 is configured to determine a maximum runable frequency of each computing resource, determine a minimum value from the determined highest frequency, determine a number of cold rooms corresponding to the minimum value according to a predetermined rule, and determine a location Whether the determined number of cold rooms is less than the number of computing resources that have been currently closed;
  • the second determining module 450 is configured to determine whether the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost if the determined number of cold rooms is less than the number of computing resources that have been turned off;
  • the resource enablement module 460 is configured to enable a closed computing resource if the minimum current cost after enabling a closed computing resource is lower than the current maximum current cost.
  • the control device for the computing resource can be implemented by a personal computer in practical applications.
  • the obtaining module 410, the determining module 420 and the closing module 430, and the submodule of the closing module 430 in the control device of the computing resource include a type determining module 430a and a resource closing module 430b, which can be used in the actual application.
  • Central Processing Unit (CPU) in the control device of the resource Digital Signal Processor (DSP), Microcontroller Unit (MCU) or programmable gate array Column (FPGA, Field-Programmable Gate Array) implementation.
  • DSP Digital Signal Processor
  • MCU Microcontroller Unit
  • FPGA Field-Programmable Gate Array
  • a control device for computing resources comprising:
  • the temperature acquisition module 502 is configured to periodically acquire the temperature of all computing resources.
  • the switching module 504 is configured to: when detecting that the temperature of a computing resource is higher than a preset value, turning off the computing resource whose temperature is higher than the preset value, and handing off the task to be processed by the closed computing resource to the closed operation The lowest temperature computing resource in the resource is processed.
  • control device for the computing resource can be implemented by a personal computer in practical applications.
  • the temperature acquisition module 502 and the switching module 504 in the control device of the computing resource may be implemented by a central processing unit (CPU) and a digital signal processor (DSP, Digital Signal Processor) in the terminal. ), Micro Control Unit (MCU) or Field-Programmable Gate Array (FPGA).
  • CPU central processing unit
  • DSP Digital Signal Processor
  • MCU Micro Control Unit
  • FPGA Field-Programmable Gate Array
  • the disclosed apparatus and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner such as: multiple units or components may be combined, or Can be integrated into another system, or some features can be ignored or not executed.
  • the coupling, or direct coupling, or communication connection of the components shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, and may be electrical, mechanical or other forms. of.
  • the units described above as separate components may or may not be physically separated, and the components displayed as the unit may or may not be physical units, that is, may be located in one place or distributed to multiple network units; Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated in one processing order.
  • each unit may be separately used as a unit, or two or more units may be integrated into one unit; the integrated unit may be implemented in the form of hardware or a hardware plus software functional unit. Formal realization.
  • the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed.
  • the foregoing storage device includes the following steps: the foregoing storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
  • ROM read-only memory
  • RAM random access memory
  • magnetic disk or an optical disk.
  • optical disk A medium that can store program code.
  • the above-described integrated unit of the present invention may be stored in a computer readable storage medium if it is implemented in the form of a software function module and sold or used as a standalone product.
  • the technical solution of the embodiments of the present invention may be embodied in the form of a software product in essence or in the form of a software product stored in a storage medium, including a plurality of instructions.
  • a computer device (which may be a personal computer, server, or network device, etc.) is caused to perform all or part of the methods described in various embodiments of the present invention.
  • the foregoing storage medium includes various media that can store program codes, such as a mobile storage device, a ROM, a RAM, a magnetic disk, or an optical disk.
  • the technical solution of the embodiment of the present invention determines the number of cold rooms corresponding to the highest frequency according to a predetermined rule by acquiring a highest frequency that the first computing resource is currently operable in the device, and determines a specific number of second computing resources. Said that a certain number of second computing resources are closed, wherein The specific amount is the same as the number of the cold rooms.
  • the second computing resource is turned off based on the characteristic that the heat is spontaneously transmitted from the high temperature object to the low temperature object, and the temperature of the second computing resource is lowered, and the heat of the first computing resource with a higher temperature is lower to the second operation of the lower temperature.
  • the resource is delivered, thereby achieving the heat dissipation of the first computing resource.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Human Computer Interaction (AREA)
  • Air Conditioning Control Device (AREA)
  • Power Sources (AREA)

Abstract

本发明实施例公开了一种运算资源控制的方法，所述方法包括：通过获取装置中的第一运算资源支持运行的最高频率，根据预定的规则，确定与所述最高频率对应的冷房数量，确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。本发明实施例同时还公开了一种运算资源控制的装置和计算机存储介质。

Description

运算资源的控制方法、装置和计算机存储介质 技术领域
本发明涉及电子技术领域，尤其涉及一种运算资源的控制方法、装置和计算机存储介质。
背景技术
随着技术的发展，用于智能移动终端(例如手机、平板电脑)的处理器的计算能力越来越强大。处理器在极小的物理封装中包含了越来越多的运算单元，这使得处理器的发热问题越来越受到关注。现有技术中，当移动终端设备内部遇到高温时，热管理器(Thermal Management)会把一些运算单元的频率降低，甚至强制关闭几乎所有运算单元来达到降温目的。如果还是无法有效降温，作为最后一道关卡，Thermal Management就会启动装置保护，强制装置关机。此做法虽能保护装置避免高温减短寿命，但是运算中的程序容易被强制中断，影响用户体验。现有的应用于智能移动终端的处理器的热管理策略存在改善空间。
发明内容
本发明实施例期望提供一种运算资源控制的方法、装置和计算机存储介质，以解决上述问题。
本发明实施例提供了一种运算资源的控制方法，所述方法包括：获取装置中的第一运算资源支持运行的最高频率；根据预定的规则，确定与所述最高频率对应的冷房数量；
确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。
在其中一个实施例中，所述根据预定的规则，确定与最高频率对应的冷房数量的步骤包括：根据预设冷房需求表，查找与最高频率匹配的预设频率点，获取冷房需求表中与所述频率点对应的冷房数量，其中，所述冷房需求表中记录了预设频率点与冷房数量的对应关系。
在其中一个实施例中，所述确定特定数量的第二运算资源，将该特定数量的第二运算资源关闭的步骤包括：从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源；按照下述之一关闭运算资源：选择并关闭特定数量的闲置运算资源；关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述特定数量；关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
在其中一个实施例中，所述确定特定数量的第二运算资源，将特定数量的第二运算资源关闭的步骤之后，还包括：定时获取各个运算资源的温度；当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由根据所述冷房数量已关闭的运算资源中温度最低的运算资源进行处理。
在其中一个实施例中，所述确定特定数量的第二运算资源，将特定数量的第二运算资源关闭的步骤之后，还包括：确定每个运算资源支持运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与所述最小值所对应的冷房数量，判断所述确定的冷房数量是否小于当前已经被关闭的运算资源的数量；若确定的冷房数量小于已经被关闭的运算资源的数量，则进一步判断启用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低；若是，则启用一个已关闭的运算资源。
本发明实施例还提供了一种运算资源的控制方法，所述方法包括：定时获取所有运算资源的温度；
当检测到特定运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要运算的处理交由已关闭的运算资源中温度最低的运算资源进行处理。
本发明实施例还提供了一种运算资源控制的装置，所述装置包括：获取模块，配置为获取装置中的第一运算资源支持运行的最高频率；确定模块，配置为根据预定的规则，确定与所述最高频率对应的冷房数量；关闭模块，配置为确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。
在其中一个实施例中，所述确定模块，还配置为根据预设冷房需求表，查找与所述最高频率匹配的预设频率点，获取冷房需求表中与所述频率点对应的冷房数量，其中，所述冷房需求表中记录了预设频率点与冷房数量的对应关系。
在其中一个实施例中，所述关闭模块包括：类型确定模块，配置为从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源；资源关闭模块，配置为按照下述之一关闭运算资源：
选择并关闭特定数量的闲置运算资源；
关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述特定数量；
关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
在其中一个实施例中，上述装置还包括：温度获取模块，配置为定时获取所有运算资源的温度；交换模块，当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
在其中一个实施例中，上述装置还包括：第一判断模块，配置为确定每个运算资源支持运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与所述最小值所对应的冷房数量，判断所述确定的冷房数量是否小于当前已经被关闭的运算资源的数量；第二判断模块，若所述确定的冷房数量小于已经被关闭的运算资源的数量，则进一步判断启用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低；资源启用模块，配置为若启用一个已关闭的运算资源后的最小电流花费比当前的最大电流花费低，则启用一个已关闭的运算资源。
本发明实施例还提供了一种运算资源的控制装置，该装置包括：
温度获取模块，配置为定时获取所有运算资源的温度；
交换模块，配置为当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
本发明实施例还提供了一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行本发明实施例所述的运算资源的控制方法。
本发明实施例提供的运算资源的控制方法、装置和计算机存储介质，通过获取装置中的第一运算资源当前可运行的最高频率，根据预定的规则，确定与所述最高频率对应的冷房数量，确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。基于热量会自发的由高温物体传递到低温物体传递的特性，关闭第 二运算资源，第二运算资源的温度降低，那么温度较高的第一运算资源的热量会向温度较低的第二运算资源传递，从而实现了第一运算资源的散热。
附图说明
图1为一个实施例中运算资源控制的方法的流程图；
图2为一个实施例中进行运算资源关闭的流程图；
图3为一个实施例中启用一个已关闭的运算资源的流程图；
图4为一个实施例中另一种运算资源控制的方法的流程图；
图5为一个实施例中运算资源控制的装置的结构框图；
图6为一个实施例中关闭模块的结构框图；
图7为另一个实施例中运算资源控制的装置的结构框图；
图8为一个实施例中另一种运算资源控制的装置的结构框图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
如图1所示，在一个实施例中，提出了一种运算资源控制的方法，所述方法可以应用于多种电子装置中，包括但不限于个人计算机、笔记本电脑、智能手机、平板电脑、智能式穿戴设备等，所述方法具体包括：
步骤110，获取装置中的第一运算资源支持运行的最高频率。
所述运算资源可以为处理器中的中央处理单元(Central Processing Unit,CPU)的核心，例如，现有典型的用于移动终端的处理包含的中央处理单元可以具有4个、6个或8个核心。所述运算资源还可以为处理器中的图形处理单元(Graphics Processing Unit,GPU)。可以理解地，本领域技术人员会意识到，所述运算资源并不局限于前述二者，在其他实施方式中，所述 运算资源可以为处理器中的核心、主核心、子核心、硬件引擎等具有计算能力的组件。上述运算资源可以为上述的单独一种，或是上述的多种之组合。
本实施例中，装置中的运算资源通常都支持多频率点运算，一般是根据运算的负载量和使用率在所有可选的频率点中选择合适的频率点作为运行频率。最高频率是指在所有可选的频率点中最大的那个频率。为了保护处理器因过热而受损，热管理器会在侦测到高温时强制降低运算资源可挑选的频率点，最高频率也随之下降。
比如某个运算资源刚开始有10个频率点可使用：634MHz、768MHz、864MHz、960MHZ、1248MHz、1344MHz、1440MHz、1536MHz、1632MHz、1766MHz，此时的最高频率为1766MHz，当运算资源温度较高时，可用频率点逐级下降，例如先降到只有9个频率点可用，此时的最高频率是1632MHz，如果温度还是持续上升，频率点就降到只有8个可用，此时的最高频率降到了1536MHz，以此类推。
步骤120，根据预定的规则，确定与所述最高频率对应的冷房数量。
本实施例中，预先设置冷房需求表，并存储在装置中，冷房需求表中记录了预设频率点与冷房数量的对应关系。具体的，查找冷房需求表中与所述最高频率匹配的预设频率点，再根据所述预设频率点确定冷房数量。在本实施例中，冷房需求表中记录的频率点是对应于不同冷房数量的频率临界点，当最高频率处在两个频率临界点之间时，与最高频率匹配的频率点是较大的那个频率点。在其他实施方式中，可以基于运算资源的最大频率通过特定的公式计算出冷房数量。
比如，在一个具体的实例中，某个处理器有四个运算资源可以使用，提供的频率点有384MHz、480MHz、768MHz、864MHz、960MHz、1248MHz、1344MHz、1632MHz、1766MHz。预设的冷房需求表可如表1所示。
1. 768MHz 2. 1248MHz 3. 1766MHz
4.冷房数量＝2 5.冷房数量＝1 6.冷房数量＝0
表1
其中，当768MHz<最高频率<1248MHz时，此时最高频率虽然小于1248MHz，但还未降到768MHz，那么与目前最高频率匹配的频率点为1248MHz，对应的冷房数量为1。当1248MHz<最高频率<1766MHz时，与目前最高频率匹配的频率点为1766MHz。当最高频率<768MHz时，与目前最高频率匹配的频率点为768MHz。
步骤130，确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与冷房数量相同。
当第一运算资源的工作负载较高但当前可运行的最高频率较低时，表明第一运算资源因温度较高而导致其可运行的最高频率被降低，此时需要进行散热。本实施例通过关闭第二运算资源来进行散热。具体的，关闭第二运算资源，即第二运算资源被断电，第二运算资源的温度随之降低。第二运算资源的温度低于第一运算资源时，考虑到热传导的方向性，即热量会自发的由高温物体传递到低温物体，温度较高的第一运算资源的热量会向温度较低的第二运算资源传递，从而实现第一运算资源的散热。被关闭的第二运算资源因此被称为冷房。
在本实施方式中，第二运算资源和第一运算资源的类别相同，均为处理器中的中央处理单元的核心。可以理解的，在其他实施方式中，在需要时，第二运算资源和第一运算资源的类别可以不相同，例如，第一运算资源为处理器中的中央处理单元的核心，第二运算资源可以为处理器中的图形处理单元、硬件引擎等具有计算能力的组件之一或者多种之组合。
从多个第二运算资源中选出特定数量的第二运算资源不是随机的，需要考虑到的众多因素中的一个是第二运算资源的物理布局，即从离第一运算资源近的第二运算资源中选择。
如图2所示，在一个实施例中，确定特定数量的第二运算资源，将特定数量的第二运算资源关闭的步骤包括：
步骤130a，从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源。
在本实施例中，预先将运算资源的类型分为三类，分别是闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源。其中，闲置运算资源是指没有运算任务，处于空闲状态的运算资源，高负载且长时间使用的运算资源是指一段时间内一直处于忙碌状态的运算资源，最不忙碌的运算资源是指运行负载最低的运算资源。
步骤130b，按照下述之一关闭运算资源：
选择并关闭特定数量的闲置运算资源；
关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述特定数量；
关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
在本实施例中，根据冷房数量，关闭与该冷房数量相等的运算资源。具体的，如果数量满足要求，优先选择关闭闲置运算资源。当闲置运算资源的数量少于冷房数量时，关闭全部闲置运算资源，再选择关闭部分高负载且长时间使用运算资源。当全部闲置运算资源和全部高负载且长时间使用的运算资源的数量少于冷房数量时，将前述二者全部关闭，再选择关闭部分最不忙碌的运算资源。
例如，假设从冷房需求表中查找到的需要的冷房数量为4个，则需要关闭4个运算资源，优先关闭闲置运算资源。若当前闲置运算资源只有1 个，那么剩下的3个再选择关闭高负载且长时间使用的运算资源。若高负载且长时间使用的运算资源的数量不足3个，余下的再选择关闭最不忙碌的运算资源。
本实施例中，优先关闭闲置运算资源是因为其没有运算任务，但为其配置的时钟源和电源还没有完全断开，只是进入了较低电流模式，将其关闭供应的电才会完全断开，这样就没有了任何电转换成热的可能性，能最优先帮助散热，所以优先关闭闲置运算资源最合适。如果闲置运算资源数量不足，接着去挑选关闭高负载且长时间使用的运算资源，是因为其高负载持续一段时间，温度会比较高，温度越高，电产生的热就会越多，通常呈指数型增长，这样导致电源必须供应越来越大的电流才能抵消在高温下电转换成热造成的浪费，而电流越大产生的热量也会越多，这样热量将会越来越多，将很不利于散热。为了避免此情况，高负载且长时间使用的运算资源作为第二优先对象去关闭。如果前述二者的数量仍不满足要求，选择关闭最不忙碌的运算资源，其温度处于居中平衡状态，没有迫切需要关闭的需求，所以留在最后。通过合理的选择关闭运算资源的顺序，能够更有效的进行散热。
在一个实施例中，为了使装置散热更加的全面均匀，还可以进行冷房轮替。在确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭的步骤130之后，还包括：
定时获取各个运算资源的温度。
本实施例中，每个运算资源附近安置有温升传感器(Thermal Sensor)，可以快速的侦测每个运算资源的温度。为了及时了解每个运算资源温度的变化情况，可以定时(比如1秒)获取温升传感器的数据以确定各个运算资源的温度。
当检测到某个运算资源的温度高于预设值时，则关闭温度高于预设值 的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
上述步骤中的运算资源均属于相同的类别，例如，均为处理器的中央处理单元中的核心。
在本实施例中，预先为各个运算资源的温度设定一个预设值，根据冷房数量进行关闭运算资源以后，如果发现运算资源的温度还是比较高，但是还不足以再关闭一个运算资源时，那么当检测到某个运算资源的温度高于预设值时，则将该运算资源进行关闭，同时把该关闭的运算资源所需要运算的任务交由当前已关闭的运算资源中温度最低的运算资源进行处理。这样通过关闭温度高的运算资源，同时启用已关闭的温度最低的运算资源，能够使散热更加的全面均匀。
在一个实施例中，为了合理运用运算资源，已关闭的运算资源可以重新启用，将其恢复成可协同运算的运算资源。如图3所示，确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭的步骤之后，还包括：
步骤142，确定每个运算资源的可运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与该最小值所对应的冷房数量，判断所述确定的冷房数量是否小于当前已经被关闭的运算资源的数量，若是，则进入步骤144，否则结束。需要注意的是，在一个实施方式中，如果当前运行的运算资源共享同一时钟源(clock source)，所述当前运行的运算资源的可运行的最高频率相同，所述最小值即是当前运行的运算资源的可运行的最高频率。
在本实施例中，在关闭特定数量的第二运算资源以后，再次确定每个运算资源的可运行的最高频率，从所有运算资源的最高频率中确定最小值，根据预先设置的冷房需求表，查找与最小值匹配的冷房需求表中的预设频 率点，再根据所述预设频率点查找冷房需求表中对应的冷房数量。冷房需求表中记录的频率点是对应于不同冷房数量的频率临界点，当最小值处在两个频率临界点之间时，与最小值匹配的频率点是较小的那个频率点。
具体的，如表1所示，其中，当768MHz<最小值<1248MHz时，此时最小值虽然大于768MHz，但是小于1248MHz，那么与最小值匹配的频率点为768MHz，对应的冷房数量为2。当1248MHz<最小值<1766MHz时，与目前最小值匹配的频率点为1248MHz，对应的冷房数量为1。当最小值>1766MHz时，与目前最小值匹配的频率点为1766MHz，对应的冷房数量为0。例如，当前的冷房数量为2，当所述最小值回到1248MHz以上才允许释放一间冷房(即启用一个已关闭的运算资源)，1248MHz是此时释放一间冷房必须达到的最小频点。
步骤144，判断启用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低，若是，则进入步骤306，否则结束。
具体的，电流花费即耗电量，其计算公式为Power(f)*M，其中f代表频率，Power(f)代表在频率f下的功耗，M代表正在运行的运算资源的数量。最小电流花费是指频率在最小频点时的电流花费，启用一个已关闭的运算资源其最小电流花费可以表示为Power(fmin)*(M+1)，其中fmin代表使用一个已关闭的运算资源可运行的最高频率必须达到的最小频点，这个最小频点从冷房需求表中查找得到，由于启用了一个已关闭的运算资源，那么正在运行的运算资源就变成了M+1个。当前最大电流花费表示为Power(F)*M，代表冷房数量M不变的情况下，运算资源执行在当前最高频率F下的电流花费，其中，F代表当前可允许的最高频率。判断使用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低，也就是判断当前的最大用电量是否足够承受的住释放一间冷房后的用电量，若是，则进行步骤146，若否，则结束。
步骤146，启用一个已关闭的运算资源。
在本实施例中，若当前的冷房数量大于所需要的冷房数量，并且启用一个已关闭的运算资源后的最小电流花费比当前的最大电流花费低，那么就释放一间冷房即启用一个已关闭的运算资源。这样可以充分合理的利用运算资源，而不至于造成运算资源的浪费。
本发明实施例还提供了一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行本发明实施例所述的运算资源的控制方法。
如图4所示，在一个实施例中，提出了一种运算资源的控制方法，该方法包括：
步骤202，定时获取所有运算资源的温度。
本实施例中，每个运算资源附近安置有温升传感器(Thermal Sensor)，可以快速的侦测每个运算资源的温度。为了及时了解每个运算资源温度的变化情况，可以定时(比如1秒)获取温升传感器的数据以确定各个运算资源的温度。
步骤204，当检测到特定运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
上述步骤202和204中的运算资源均属于相同的类别，例如，均为处理器的中央处理单元中的核心。
在本实施例中，预先为各个运算资源的温度设定一个预设值，根据冷房数量进行关闭运算资源以后，如果发现运算资源的温度还是比较高，但是还不足以再关闭一个运算资源时，那么当检测到某个运算资源的温度高于预设值时，则将该运算资源进行关闭，同时把该关闭的运算资源所需要运算的任务交由当前已关闭的运算资源中温度最低的运算资源进行处理。 这样通过关闭温度高的运算资源，同时启用已关闭的温度最低的运算资源，能够使散热更加的全面均匀。
在一个实施方式中，上述步骤202和204设置在步骤130之后。
本发明实施例还提供了一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行本发明实施例所述的运算资源的控制方法。
如图5所示，在一个实施例中，提出了一种运算资源的控制装置，该装置包括：
获取模块410，配置为获取装置中的第一运算资源支持运行的最高频率。
确定模块420，配置为根据预定的规则，确定与最高频率对应的冷房数量。
关闭模块430，配置为确定特定数量的第二运算资源，将特定数量的第二运算资源关闭，其中，特定数量与冷房数量相同。
在一个实施例中，所述确定模块420，还配置为根据预设冷房需求表，查找与所述最高频率匹配的预设频率点，获取冷房需求表中与所述频率点对应的冷房数量，其中，所述冷房需求表中记录了预设频率点与冷房数量的对应关系。
如图6所示，在一个实施例中，所述关闭模块430包括：
类型确定模块430a，配置为从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源。
资源关闭模块430b，配置为按照下述之一关闭运算资源：
选择并关闭特定数量的闲置运算资源；
关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述 特定数量；
关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
在一个实施例中，上述装置还包括：
温度获取模块，配置为定时获取各个运算资源的温度。
交换模块，配置为当检测到某个运算资源的温度高于预设值时，则关闭该温度高于预设值的运算资源，将该关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
如图7所示，在一个实施例中，上述装置还包括：
第一判断模块440，配置为确定每个运算资源的可运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与所述最小值所对应的冷房数量，判断所述确定的冷房数量是否小于当前已经被关闭的运算资源的数量；
第二判断模块450，配置为若确定的冷房数量小于已经被关闭的运算资源的数量，则进一步判断启用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低；
资源启用模块460，配置为若启用一个已关闭的运算资源后的最小电流花费比当前的最大电流花费低，则启用一个已关闭的运算资源。
本发明实施例中，所述运算资源的控制装置在实际应用中，可通过个人计算机实现。所述运算资源的控制装置中的获取模块410、确定模块420和关闭模块430，以及所述关闭模块430的子模块包括类型确定模块430a和资源关闭模块430b，在实际应用中均可由所述运算资源的控制装置中的中央处理器(CPU，Central Processing Unit)、数字信号处理器(DSP，Digital Signal Processor)、微控制单元(MCU，Microcontroller Unit)或可编程门阵 列(FPGA，Field－Programmable Gate Array)实现。
如图8所示，在一个实施例中，提出了一种运算资源的控制装置，所述装置包括：
温度获取模块502，配置为定时获取所有运算资源的温度。
交换模块504，配置为当检测到某个运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
本发明实施例中，所述运算资源的控制装置在实际应用中，可通过个人计算机实现。所述运算资源的控制装置中的温度获取模块502和交换模块504，在实际应用中均可由所述终端中的中央处理器(CPU，Central Processing Unit)、数字信号处理器(DSP，Digital Signal Processor)、微控制单元(MCU，Microcontroller Unit)或可编程门阵列(FPGA，Field－Programmable Gate Array)实现。
在本申请所提供的几个实施例中，应该理解到，所揭露的设备和方法，可以通过其它的方式实现。以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，如：多个单元或组件可以结合，或可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的各组成部分相互之间的耦合、或直接耦合、或通信连接可以是通过一些接口，设备或单元的间接耦合或通信连接，可以是电性的、机械的或其它形式的。
上述作为分离部件说明的单元可以是、或也可以不是物理上分开的，作为单元显示的部件可以是、或也可以不是物理单元，即可以位于一个地方，也可以分布到多个网络单元上；可以根据实际的需要选择其中的部分或全部单元来实现本实施例方案的目的。
另外，在本发明各实施例中的各功能单元可以全部集成在一个处理单 元中，也可以是各单元分别单独作为一个单元，也可以两个或两个以上单元集成在一个单元中；上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
本领域普通技术人员可以理解：实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成，前述的程序可以存储于一计算机可读取存储介质中，该程序在执行时，执行包括上述方法实施例的步骤；而前述的存储介质包括：移动存储设备、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
或者，本发明上述集成的单元如果以软件功能模块的形式实现并作为独立的产品销售或使用时，也可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明实施例的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机、服务器、或者网络设备等)执行本发明各个实施例所述方法的全部或部分。而前述的存储介质包括：移动存储设备、ROM、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以所述权利要求的保护范围为准。
工业实用性
本发明实施例的技术方案通过获取装置中的第一运算资源当前可运行的最高频率，根据预定的规则，确定与所述最高频率对应的冷房数量，确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中， 所述特定数量与所述冷房数量相同。基于热量会自发的由高温物体传递到低温物体传递的特性，关闭第二运算资源，第二运算资源的温度降低，那么温度较高的第一运算资源的热量会向温度较低的第二运算资源传递，从而实现了第一运算资源的散热。

Claims (14)

  1. 一种运算资源的控制方法，所述方法包括：
    获取装置中的第一运算资源支持运行的最高频率；
    根据预定的规则，确定与所述最高频率对应的冷房数量；
    确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。
  2. 根据权利要求1所述的方法，其中，所述根据预定的规则，确定与所述最高频率对应的冷房数量的步骤包括：
    根据预设冷房需求表，查找与所述最高频率匹配的预设频率点，获取冷房需求表中与所述频率点对应的冷房数量，其中，所述冷房需求表中记录了预设频率点与冷房数量的对应关系。
  3. 根据权利要求1所述的方法，其中，所述确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭的步骤包括：
    从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源；
    按照下述之一关闭运算资源：
    选择并关闭特定数量的闲置运算资源；
    关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述特定数量；
    关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
  4. 根据权利要求1所述的方法，其中，所述确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭的步骤之后，还包括：
    定时获取所有运算资源的温度；
    当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由根据所述冷房数量已关闭的运算资源中温度最低的运算资源进行处理。
  5. 根据权利要求1所述的方法，其中，所述确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭的步骤之后，还包括：
    确定每个运算资源支持运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与所述最小值所对应的冷房数量，判断所述确定的冷房数量是否小于已经被关闭的运算资源的数量；
    若所述确定的冷房数量小于已经被关闭的运算资源的数量，则进一步判断启用一个已关闭的运算资源后的最小电流花费是否比当前的最大电流花费低；若是，则启用一个已关闭的运算资源。
  6. 一种运算资源的控制方法，所述方法包括：
    定时获取所有运算资源的温度；
    当检测到特定的运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
  7. 一种运算资源控制装置，所述装置包括：
    获取模块，配置为获取装置中的第一运算资源支持运行的最高频率；
    确定模块，配置为根据预定的规则，确定与所述最高频率对应的冷房数量；
    关闭模块，配置为确定特定数量的第二运算资源，将所述特定数量的第二运算资源关闭，其中，所述特定数量与所述冷房数量相同。
  8. 根据权利要求7所述的装置，其中，所述确定模块，还配置为根据预设冷房需求表，查找与所述最高频率匹配的预设频率点，获取冷房需求 表中与所述频率点对应的冷房数量，其中，所述冷房需求表中记录了预设频率点与冷房数量的对应关系。
  9. 根据权利要求7所述的装置，其中，所述关闭模块包括：
    类型确定模块，配置为从多个第二运算资源中确定闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源；
    资源关闭模块，配置为按照下述之一关闭运算资源：
    选择并关闭特定数量的闲置运算资源；
    关闭全部的闲置运算资源，以及部分高负载且长时间使用的运算资源，其中闲置运算资源和高负载且长时间使用的运算资源的数量之和等于所述特定数量；
    关闭全部闲置运算资源和全部高负载且长时间使用的运算资源，以及部分最不忙碌的运算资源，闲置运算资源、高负载且长时间使用的运算资源和最不忙碌的运算资源的数量之和等于所述特定数量。
  10. 根据权利要求7所述的装置，其中，所述装置还包括：
    温度获取模块，配置为定时获取所有运算资源的温度；
    交换模块，配置为当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由根据所述冷房数量已关闭的运算资源中温度最低的运算资源进行处理。
  11. 根据权利要求7所述的装置，其中，所述装置还包括：
    第一判断模块，配置为确定每个运算资源支持运行的最高频率，从确定的所述最高频率中确定最小值，按照预定的规则确定与所述最小值所对应的冷房数量，判断所述确定的冷房数量是否小于当前已经被关闭的运算资源的数量；
    第二判断模块，配置为若所述确定的冷房数量小于已经被关闭的运算资源的数量，则进一步判断启用一个已关闭的运算资源后的最小电流花费 是否比当前的最大电流花费低；
    资源启用模块，配置为若启用一个已关闭的运算资源后的最小电流花费比当前的最大电流花费低，则启用一个已关闭的运算资源。
  12. 一种运算资源控制装置，所述装置包括：
    温度获取模块，配置为定时获取所有运算资源的温度；
    交换模块，配置为当检测到运算资源的温度高于预设值时，则关闭温度高于预设值的运算资源，将关闭的运算资源所需要处理的任务交由已关闭的运算资源中温度最低的运算资源进行处理。
  13. 一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行权利要求1至5任一项所述的运算资源的控制方法。
  14. 一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行权利要求6所述的运算资源的控制方法。
PCT/CN2016/087227 2015-07-28 2016-06-27 运算资源的控制方法、装置和计算机存储介质 WO2017016359A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US15/747,482 US10621011B2 (en) 2015-07-28 2016-06-27 Method and apparatus for controlling operation resources, and computer storage medium
EP16829725.7A EP3330834B1 (en) 2015-07-28 2016-06-27 Method and apparatus for controlling operation resources, and computer storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510451706.4 2015-07-28
CN201510451706.4A CN105068871B (zh) 2015-07-28 2015-07-28 运算资源的控制方法和装置

Publications (1)

Publication Number Publication Date
WO2017016359A1 true WO2017016359A1 (zh) 2017-02-02

Family

ID=54498248

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/087227 WO2017016359A1 (zh) 2015-07-28 2016-06-27 运算资源的控制方法、装置和计算机存储介质

Country Status (4)

Country Link
US (1) US10621011B2 (zh)
EP (1) EP3330834B1 (zh)
CN (1) CN105068871B (zh)
WO (1) WO2017016359A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112988739A (zh) * 2019-12-18 2021-06-18 阿里巴巴集团控股有限公司 数据管理和处理方法、装置、计算机系统及可读存储介质

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105068871B (zh) 2015-07-28 2018-12-21 深圳市万普拉斯科技有限公司 运算资源的控制方法和装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101111814A (zh) * 2004-12-30 2008-01-23 英特尔公司 多核架构中的工作点管理
CN101114187A (zh) * 2006-07-26 2008-01-30 国际商业机器公司 用于监视和控制多核心处理器中的发热的方法和设备
CN103226462A (zh) * 2012-01-31 2013-07-31 Lg电子株式会社 移动终端及其控制方法以及其记录介质
CN105068871A (zh) * 2015-07-28 2015-11-18 深圳市万普拉斯科技有限公司 运算资源的控制方法和装置

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7032119B2 (en) * 2000-09-27 2006-04-18 Amphus, Inc. Dynamic power and workload management for multi-server system
US7028196B2 (en) * 2002-12-13 2006-04-11 Hewlett-Packard Development Company, L.P. System, method and apparatus for conserving power consumed by a system having a processor integrated circuit
US7421623B2 (en) * 2004-07-08 2008-09-02 International Business Machines Corporation Systems, methods, and media for controlling temperature in a computer system
US7263457B2 (en) * 2006-01-03 2007-08-28 Advanced Micro Devices, Inc. System and method for operating components of an integrated circuit at independent frequencies and/or voltages
US8589931B2 (en) * 2009-03-18 2013-11-19 International Business Machines Corporation Environment based node selection for work scheduling in a parallel computing system
US9292662B2 (en) * 2009-12-17 2016-03-22 International Business Machines Corporation Method of exploiting spare processors to reduce energy consumption
US8996330B2 (en) * 2011-01-06 2015-03-31 Qualcomm Incorporated Method and system for managing thermal policies of a portable computing device
CN102841808B (zh) * 2011-06-21 2017-12-08 技嘉科技股份有限公司 电脑系统的效能提升方法及其电脑系统
US10013511B2 (en) * 2012-04-09 2018-07-03 Purdue Research Foundation System and method for energy usage accounting in software applications
US8959371B2 (en) * 2012-07-17 2015-02-17 Freescale Semiconductor, Inc. Techniques for reducing processor power consumption through dynamic processor resource allocation
US9075611B2 (en) * 2012-09-12 2015-07-07 Htc Corporation Electronic device with power management mechanism and power management method thereof
CN104424156A (zh) * 2013-09-09 2015-03-18 中兴通讯股份有限公司 处理器的核处理方法、装置及终端
US9619289B2 (en) * 2014-09-11 2017-04-11 Dell Products, L.P. Workload optimized server for intelligent algorithm trading platforms

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101111814A (zh) * 2004-12-30 2008-01-23 英特尔公司 多核架构中的工作点管理
CN101114187A (zh) * 2006-07-26 2008-01-30 国际商业机器公司 用于监视和控制多核心处理器中的发热的方法和设备
CN103226462A (zh) * 2012-01-31 2013-07-31 Lg电子株式会社 移动终端及其控制方法以及其记录介质
CN105068871A (zh) * 2015-07-28 2015-11-18 深圳市万普拉斯科技有限公司 运算资源的控制方法和装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3330834A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112988739A (zh) * 2019-12-18 2021-06-18 阿里巴巴集团控股有限公司 数据管理和处理方法、装置、计算机系统及可读存储介质
CN112988739B (zh) * 2019-12-18 2024-04-19 阿里巴巴集团控股有限公司 数据管理和处理方法、装置、计算机系统及可读存储介质

Also Published As

Publication number Publication date
CN105068871A (zh) 2015-11-18
US20180217877A1 (en) 2018-08-02
EP3330834B1 (en) 2022-03-02
EP3330834A1 (en) 2018-06-06
CN105068871B (zh) 2018-12-21
EP3330834A4 (en) 2018-07-25
US10621011B2 (en) 2020-04-14

Similar Documents

Publication Publication Date Title
JP6873082B2 (ja) 熱アウェアデバイスブーティングのためのシステムおよび方法
KR101534450B1 (ko) 누설 전류 측정으로부터 열 관리 폴리시를 결정하는 시스템 및 방법
US9235243B2 (en) Thermal profile optimization techniques
JP6218967B2 (ja) マルチプロセッサシステムオンチップにおけるエネルギー効率認識型熱管理
US8195970B2 (en) System and method for adapting a power usage of a server during a data center cooling failure
EP3030946B1 (en) Intelligent multicore control for optimal performance per watt
CN108780349B (zh) 用于在具有异构集群架构的片上系统中进行智能热管理的系统和方法
US10394293B2 (en) Method for preventing over-heating of a device within a data processing system
US9552046B2 (en) Performance management methods for electronic devices with multiple central processing units
US8634952B2 (en) Fan control method and medium storing fan control program
KR20130105899A (ko) 휴대용 컴퓨팅 디바이스의 열 정책들을 관리하는 방법 및 시스템
JP2016527630A (ja) チップ上のマルチプロセッサシステムにおけるアイドル状態最適化のためのシステムおよび方法
CN102880477B (zh) 一种实现计算机启动的方法和计算机
US20120323400A1 (en) Optimized temperature-driven device cooling
JP2017502383A (ja) ポータブルコンピューティングデバイスにおけるシステムオンチップのマルチ相関学習型熱管理のためのシステムおよび方法
JP6240225B2 (ja) ポータブルコンピューティングデバイスにおける電圧モードの温度駆動型選択のためのシステムおよび方法
WO2017016359A1 (zh) 运算资源的控制方法、装置和计算机存储介质
EP3330853B1 (en) Method and system for controlling operation unit, and computer storage medium
US10488900B2 (en) Heat dissipation control method for operation resources, operation control system, and storage medium
CN114356570A (zh) 一种处理器调节方法、系统及电子设备
CN116914326A (zh) 电池温度控制方法、装置、终端及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16829725

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 15747482

Country of ref document: US