WO2020238326A1 - 处理器的性能状态的调节方法、装置、uefi及存储介质 - Google Patents

处理器的性能状态的调节方法、装置、uefi及存储介质 Download PDF

Info

Publication number
WO2020238326A1
WO2020238326A1 PCT/CN2020/079022 CN2020079022W WO2020238326A1 WO 2020238326 A1 WO2020238326 A1 WO 2020238326A1 CN 2020079022 W CN2020079022 W CN 2020079022W WO 2020238326 A1 WO2020238326 A1 WO 2020238326A1
Authority
WO
WIPO (PCT)
Prior art keywords
processor
performance state
state value
current
busy ratio
Prior art date
Application number
PCT/CN2020/079022
Other languages
English (en)
French (fr)
Inventor
张涛
李德辉
戴慧敏
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2020238326A1 publication Critical patent/WO2020238326A1/zh

Links

Images

Classifications

    • 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
    • G06F1/3234—Power saving characterised by the action undertaken
    • G06F1/324—Power saving characterised by the action undertaken by lowering clock frequency
    • 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
    • G06F1/04—Generating or distributing clock signals or signals derived directly therefrom
    • G06F1/08—Clock generators with changeable or programmable clock frequency

Definitions

  • the present disclosure relates to the field of communications, and in particular to a method, device, UEFI, storage medium, and electronic device for adjusting the performance state of a processor.
  • the processor power state (CPU Power State, referred to as C-State) is the power state defined by the advanced configuration and power management interface (Advanced Configuration and Power Management Interface, referred to as ACPI) specifications when the processor is working at G0. These power supplies
  • the state includes C0, C1, C2, C3...Cx, Intel Intel processor can support up to C7, from C0 to Cx processor sleep state in turn deepens, the energy consumed will also be reduced in turn.
  • the ACPI specification also defines the performance state.
  • the Intel platform usually refers to the enhanced Intel SpeedStep Technology (EIST) Operating points. Mainly use the voltage/performance status level to control the energy consumed in the C0 state.
  • the main factors of the performance status are voltage and performance status levels: higher performance status levels require higher voltage support, and this switching can occur without system reset.
  • the processor performance state level the lower the energy consumption.
  • Traditional processor energy-saving solutions are usually completed by the power supply strategy of the operating system or the central processing unit (CPU) frequency modulation strategy, and energy saving is achieved by controlling the processor C-State and performance state level.
  • the operating system When the operating system is running, it will adjust the processor performance state level according to the current processor load status to reduce the processor energy under the premise of meeting current usage.
  • the Windows operating system configures the operating system energy saving by setting the power plan
  • the Linux operating system can control the processor performance status level to achieve energy saving by configuring the CPU frequency modulation strategy, but the performance status adjustment of the operating system often changes too much in terms of power saving.
  • the performance is only a general level, and it is not optimal in terms of performance and energy saving.
  • Traditional energy saving methods have poor performance in precision control and energy efficiency utilization.
  • the embodiments of the present disclosure provide a method, device, UEFI, storage medium, and electronic device for adjusting the performance state of a processor, so as to at least solve the problem of low accuracy of adjusting the performance state of the processor in related technologies, which may easily lead to energy efficiency utilization.
  • the problem of low rate may be used to determine the performance state of a processor's performance state of a processor's performance state of a processor's performance state of a processor'sonic device.
  • a method for adjusting the performance state of a processor including: obtaining a current busy ratio of the processor, wherein the current busy ratio is the current resource increment value of the processor core CORE The ratio to the current performance state value of the processor; the target performance state value corresponding to the current busy ratio is determined according to a pre-generated configuration file, wherein the configuration file records the adjustment granularity of the busy ratio and the performance state value ; Adjust the performance state value of the processor to the target performance state value.
  • an apparatus for adjusting the performance state of a processor including: an obtaining module configured to obtain the current busy ratio of the processor, wherein the current busy ratio is the processor core The ratio of the current CORE resource increment value to the current performance state value of the processor; the processing module is configured to determine the target performance state value corresponding to the current busy ratio according to a pre-generated configuration file, wherein the configuration file The adjustment granularity of the busy ratio and the performance state value is recorded; the adjustment module is configured to adjust the performance state value of the processor to the target performance state value.
  • a unified and extensible firmware interface UEFI is provided, which includes the adjustment device described in the foregoing embodiment.
  • a storage medium in which a computer program is stored, wherein the computer program is configured to execute the steps in any one of the above method embodiments when running.
  • an electronic device including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the foregoing Steps in the method embodiment.
  • the target performance status value of the processor can be determined according to the busy ratio and utilization of the processor, and the performance status value of the processor can be adjusted to the target performance status value, thereby realizing timely
  • the performance state is adjusted to the performance state value corresponding to the current busy ratio and utilization rate, which effectively avoids the problem of excessive adjustment and waste of energy efficiency, and achieves the effect of improving the adjustment accuracy of the performance state of the processor and improving the energy efficiency utilization rate.
  • Fig. 1 is a flowchart of a method for adjusting the performance state of a processor according to an embodiment of the present disclosure
  • FIG. 2 is a flowchart of a method for UEFI firmware to control the performance state of a processor according to an embodiment of the present disclosure
  • Figure 3 is a flowchart of a method for UEFI firmware to adjust performance status according to an embodiment of the present disclosure
  • FIG. 4 is a flowchart of a method for real-time adjustment of processor performance status by UEFI firmware according to an embodiment of the present disclosure
  • FIG. 5 is a flowchart of operations performed by a processor busy ratio acquisition module according to an embodiment of the present disclosure
  • Fig. 6 is a schematic structural diagram of a processor performance state adjustment module according to an embodiment of the present disclosure
  • Fig. 7 is a flowchart of operations performed by a processor performance state adjustment module according to an embodiment of the present disclosure
  • Fig. 8 is a structural block diagram of a device for adjusting the performance state of a processor according to an embodiment of the present disclosure.
  • FIG. 1 is a flowchart of the method for adjusting the performance state of a processor according to an embodiment of the present disclosure. As shown in FIG. 1, the process includes the following steps :
  • Step S102 Obtain the current busy ratio of the processor, where the current busy ratio is the ratio of the current resource increment value of the processor core CORE to the current performance state value of the processor;
  • Step S104 Determine a target performance state value corresponding to the current busy ratio according to a pre-generated configuration file, wherein the configuration file records the adjustment granularity of the busy ratio and the performance state value;
  • Step S106 Adjust the performance state value of the processor to the target performance state value.
  • the foregoing operations may be performed by a unified extensible firmware interface (Unified Extensible Firmware Interface, referred to as UEFI) or similar functional module management independent of the OS system or software under the system.
  • UEFI Unified Extensible Firmware Interface
  • the target performance state value can be based on the busy ratio of the processor, based on the utilization conservation criterion, and combined with the different energy efficiency modes currently used and analyzed according to the application pressure to achieve precise control of the processor performance state.
  • Energy efficiency modes can be divided into different modes according to energy saving and performance requirements.
  • the core resource increment value of the processor at the next moment is the same as the current moment, that is, the processor core resource increment value remains unchanged.
  • the P_Current required by BUSY_x can be calculated. The larger the value of BUSY_x, the higher the efficiency of processor resource utilization. Among them, the essence of conservation of utilization is that under the premise of clear operating system OS resource requirements (utilization requirements), increasing the processor frequency will cause the busy ratio of UEFI computing to decrease, and reducing the processor frequency will increase the busy ratio.
  • BUSY_x Based on the conservation of USAGE_x utilization rate, BUSY_x can be calculated, and then the required performance status value can be calculated.
  • the pre-generated configuration file may be the adjustment granularity of each busy ratio and each target performance state value obtained through multiple experimental tests in advance, wherein the adjustment granularity of the performance state value may be a positive adjustment granularity, It can also be a negative adjustment granularity, and the adjustment granularity of the performance state value can also be replaced with an adjustment algorithm.
  • the adjustment algorithm can be a positive adjustment algorithm or a negative adjustment algorithm.
  • the target performance status value of the processor can be determined according to the busy ratio of the processor, and the performance status value of the processor can be adjusted to the target performance status value, so that the performance of the processor can be adjusted in time.
  • the state is adjusted to the performance state value corresponding to the current busy ratio, which effectively avoids the problem of excessive adjustment range and energy efficiency waste, and achieves the effect of improving the adjustment accuracy of the performance state of the processor and improving the energy efficiency utilization.
  • the above steps S102-106 may be executed periodically, that is, the performance state value of the processor is adjusted once in a cycle (or called a polling cycle).
  • obtaining the current busy ratio of the processor includes: obtaining the current busy ratio of the processor through a unified extensible firmware interface UEFI; and determining the current busy ratio according to a pre-generated configuration file.
  • the target performance state value corresponding to the ratio includes: determining, by the UEFI, the target performance state value corresponding to the current busy ratio according to the pre-generated configuration file; and adjusting the performance state value of the processor to the
  • the target performance state value includes: adjusting the performance state value of the processor to the target performance state value by the UEFI.
  • server efficiency management can be implemented based on UEFI firmware, without relying on operating systems and third-party software and hardware tools, and real-time control of the server processor performance state can be achieved through UEFI firmware.
  • UEFI is based on periodic polling to monitor the CORE increment value of the processor in real time and calculate the busy ratio. Then, according to the busy ratio, combined with the pre-generated configuration file, it can achieve precise control of the processor performance state and avoid the waste of processor frequency resources. At the same time, based on the busy ratio and utilization rate calculated by the UEFI firmware, it can directly determine whether the OS needs to respond quickly during the polling period, so as to directly set the required performance state to maximize the business performance requirements and reduce the processor performance state step Business delays caused by slowness, and when business delays are required, the conservation of utilization can be achieved to accurately obtain the required performance status. In addition, it should be noted that the execution of the above operations by UEFI is only an optional implementation, and it is not limited to this in practical applications.
  • determining the target performance state value corresponding to the current busy ratio according to a pre-generated configuration file includes: determining a pre-created performance state level support range table of the processor, wherein the The performance state level support range table records at least the size of the main frequency supported by the processor divided according to predetermined frequency intervals, and the minimum main frequency recorded in the performance state level support range table is the minimum supported by the processor The main frequency, the maximum main frequency recorded in the performance state level support range table is the maximum main frequency supported by the processor; the adjustment of the current performance state value corresponding to the current busy ratio is determined according to a pre-generated configuration file Granularity; find the main frequency value corresponding to the adjustment granularity of the current performance state value in the performance state level support range table, and use the found main frequency value as the main frequency value included in the target performance state value Frequency value.
  • the supported performance state level range table can be dynamically created according to the processor type. For example, the minimum frequency of the processor is 8 and the maximum frequency is 20.
  • the created performance state level support range table can be: ⁇ 8 ,9,10,11,12,13,14,15,16,17,18,19,20 ⁇ .
  • the predetermined frequency can be 8, or 9, or other values in the table.
  • the method before adjusting the performance state value of the processor to the target performance state value, the method further includes: when it is determined that the busy ratio exceeds a predetermined threshold, performing the processing The performance state value of the processor is adjusted to the maximum performance state value supported by the processor.
  • the busy ratio of the processor is too high, for example, when it exceeds 99%, by adjusting the performance state of the processor to the maximum level supported by the processor, it is possible to ensure that the performance of the processor is not lost. The degree of response to operating system business.
  • the method before obtaining the busy ratio of the processor, the method further includes: determining that the intelligent function management IEM function is enabled; and turning off the autonomous frequency modulation capability of the processor.
  • the performance state value of the processor includes the main frequency value of the processor and the voltage value of the processor.
  • the performance state value of the processor may also include other values for indicating the performance state of the processor.
  • the embodiment of the present disclosure provides a method for UEFI firmware (corresponding to the aforementioned UEFI) to control the performance state of the processor to achieve the effect of energy saving and efficiency improvement of the server, and the process is shown in FIG. 2.
  • this embodiment turns off the operating system's function of controlling the performance status, and completely transfers the processor performance status control to the UEFI firmware.
  • Step 1 Judge whether to enable the IEM function during the startup of the UEFI firmware. If it is enabled, proceed to the following steps.
  • Step 2 The UEFI firmware turns off the operating system's ability to control processor performance status by changing the processor and operating system power management capabilities.
  • Step 3 Turn off the autonomous frequency modulation function of the processor.
  • Step 4. Turn on the periodic polling function.
  • Step 5 Register related processing functions.
  • UEFI firmware processing function directly manipulates processor registers to set its performance status.
  • the embodiment of the present disclosure provides a method for UEFI firmware to adjust the performance state to achieve the effect of energy saving and efficiency improvement of the server, and the process is shown in FIG.
  • Step 1 Dynamically create a range table of performance status levels supported by the processor according to the type of processor.
  • the minimum frequency of the processor is 8, the maximum frequency is 20, the created performance state level support range table is: ⁇ 8,9,10,11,12,13,14,15,16,17,18,19 ,20 ⁇ ;
  • Step 2 Calculate the busy ratio under the current processor performance state.
  • Step 3 Determine whether the busy ratio exceeds the maximum threshold, if it is greater than the maximum threshold, proceed to step 4, otherwise proceed to step 5.
  • the processor performance status level is directly adjusted to the maximum value in the table.
  • Step 4 Adjust the processor performance state to the maximum level in the support range table to ensure that the processor performance is not lost, and to respond to the operating system business to the greatest extent.
  • Step 5 Calculate the required performance state value through the principle of conservation of processor utilization. At the same time, the busy ratio and utilization rate under different application pressures are collected for a long time, and the data sampling and training results under different application pressures are generated in advance Different profiles (configuration files) are configured into the corresponding energy efficiency modes, and combined with the Profile, the optimal corresponding performance state value under the current busy ratio is calculated.
  • Step 6 Look up the performance state value in the table corresponding to the optimal value in the table, and then directly adjust the processor performance state value to the value in the table.
  • the embodiment of the present disclosure provides a device for real-time adjustment of processor performance status through UEFI firmware to achieve energy-saving and efficiency-increasing effects of the server.
  • the process is shown in Figure 4:
  • the processor performance status control authority After the server enters the operating system, the processor performance status control authority has been completely handed over to the UEFI firmware, so the UEFI firmware must poll in real time to obtain the current processor busy ratio, and then dynamically adjust the processor performance status according to the busy ratio.
  • the device can reduce the energy consumed by the processor.
  • Step 1 After the server enters the operating system, UEFI will periodically poll the processor status;
  • Step 2 Determine whether the performance status needs to be adjusted. If adjustment is required, go to step 3, that is, execute the processing function registered by the UEFI firmware. The processing function mainly executes steps 3 and 4, otherwise, go to step 1.
  • Step 3 Execute the processor busy ratio acquisition module (same as the subsequent acquisition module 82), and the operation performed by this module is shown in FIG. 5.
  • Obtaining the busy ratio of the processor is used to determine the processor occupancy rate. The larger the busy ratio, the higher the current processor occupancy rate.
  • the performance state adjustment module needs to be based in turn.
  • Step 4 Execute the processor performance state adjustment module (same as the subsequent processing module 84 and adjustment module 86), and the specific units included in the module are shown in FIG. 6.
  • the busy ratio comparing unit is used to judge the processor occupancy rate in the current performance state.
  • the business model analysis unit collects the busy ratio of each business stress model for a long time, and evaluates the business stress state, as shown in Figure 7.
  • the performance state calculation unit the required performance state value of the processor under the current business pressure is accurately calculated to realize intelligent energy-saving adaptation.
  • the performance state adjustment unit completes the performance state adjustment in a polling cycle.
  • the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, of course, it can also be implemented by hardware, but in many cases the former is Better implementation.
  • the technical solution of the present disclosure essentially or the part that contributes to the existing technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium (such as ROM/RAM, magnetic disk, The optical disc) includes several instructions to enable a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the method described in each embodiment of the present disclosure.
  • a device for adjusting the performance state of the processor is also provided.
  • the device is used to implement the above-mentioned embodiments and preferred implementations, and what has been described will not be repeated.
  • the term "module" can implement a combination of software and/or hardware with predetermined functions.
  • the devices described in the following embodiments are preferably implemented by software, hardware or a combination of software and hardware is also possible and conceived.
  • Fig. 8 is a structural block diagram of a device for adjusting the performance state of a processor according to an embodiment of the present disclosure. As shown in Fig. 8, the device includes:
  • the obtaining module 82 is configured to obtain the current busy ratio of the processor, where the current busy ratio is the ratio of the current resource increment value of the processor core CORE to the current performance state value of the processor;
  • the processing module 84 is configured to determine a target performance state value corresponding to the current busy ratio according to a pre-generated configuration file, wherein the configuration file records the adjustment granularity of the busy ratio and the performance state value;
  • the adjustment module 86 is configured to adjust the performance state value of the processor to the target performance state value.
  • the processing module 84 may determine the target performance state value by determining a pre-created performance state level support range table of the processor, wherein the performance state level supports The range table records at least the size of the main frequency supported by the processor divided according to predetermined frequency intervals, the minimum main frequency recorded in the performance state level support range table is the minimum main frequency supported by the processor, The maximum main frequency recorded in the performance state level support range table is the maximum main frequency supported by the processor; the adjustment granularity of the current performance state value corresponding to the current busy ratio is determined according to a pre-generated configuration file; The performance state level support range table searches for the main frequency value corresponding to the adjustment granularity of the current performance state value, and uses the found main frequency value as the main frequency value included in the target performance state value.
  • the device is further configured to perform processing when it is determined that the busy ratio exceeds a predetermined threshold before adjusting the performance state value of the processor to the target performance state value.
  • the performance state value of the processor is adjusted to the maximum performance state value supported by the processor.
  • the device is further configured to determine that the intelligent function management IEM function is turned on before acquiring the busy ratio of the processor; and turn off the autonomous frequency modulation capability of the processor.
  • the performance state value of the processor includes the main frequency value of the processor and the voltage value of the processor.
  • a unified extensible firmware interface UEFI including the adjustment device described in any of the above embodiments.
  • each of the above modules can be implemented by software or hardware.
  • it can be implemented in the following manner, but not limited to this: the above modules are all located in the same processor; or, the above modules are combined in any combination The forms are located in different processors.
  • the embodiment of the present disclosure also provides a storage medium in which a computer program is stored, wherein the computer program is configured to execute the steps in any of the foregoing method embodiments when running.
  • the foregoing storage medium may include, but is not limited to: U disk, Read-Only Memory (Read-Only Memory, ROM for short), Random Access Memory (Random Access Memory, RAM for short), Various media that can store computer programs, such as mobile hard disks, magnetic disks, or optical disks.
  • An embodiment of the present disclosure also provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any of the foregoing method embodiments.
  • the aforementioned electronic device may further include a transmission device and an input-output device, wherein the transmission device is connected to the aforementioned processor, and the input-output device is connected to the aforementioned processor.
  • modules or steps of the present disclosure can be implemented by a general computing device, and they can be concentrated on a single computing device or distributed in a network composed of multiple computing devices.
  • they can be implemented with program codes executable by the computing device, so that they can be stored in the storage device for execution by the computing device, and in some cases, can be executed in a different order than here.
  • the method, device, UEFI, storage medium, and electronic device for adjusting the performance state of the processor provided by the embodiments of the present invention have the following beneficial effects: it solves the problem of excessive adjustment in related technologies and waste of energy efficiency.
  • the problem is to achieve the effect of improving the adjustment accuracy of the performance state of the processor and improving the energy efficiency utilization.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Power Sources (AREA)

Abstract

一种处理器的性能状态的调节方法、装置、UEFI、存储介质及电子装置，其中，该方法包括：获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值（S102）；依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度（S104）；将所述处理器的性能状态值调整为所述目标性能状态值（S106）。所述方法、装置、UEFI、存储介质及电子装置解决了相关技术中存在的调节幅度过大，能效浪费的问题，达到提高处理器的性能状态的调节精度以及提高能效利用率的效果。

Description

处理器的性能状态的调节方法、装置、UEFI及存储介质 技术领域
本公开涉及通信领域，具体而言，涉及一种处理器的性能状态的调节方法、装置、UEFI、存储介质及电子装置。
背景技术
处理器电源状态(CPU Power State，简称为C-State)，是高级配置和电源管理接口(Advanced Configuration and Power Management Interface，简称为ACPI)规范定义的处理器工作在G0时的电源状态，这些电源状态包括C0,C1,C2,C3…Cx，英特尔Intel处理器最大可以支持到C7,从C0到Cx处理器休眠状态依次加深，消耗的能量也会依次减少。
在C0态下ACPI规范还定义了性能状态，Intel平台通常指的是智能降频技术运行点Enhanced Intel SpeedStep Technology(EIST)Operating points。主要是用电压/性能状态等级来控制C0状态下消耗的能量。性能状态主要因素是电压和性能状态等级：更高的性能状态等级需要更高的电压支持，这种切换不需要系统复位即可发生。在C0态下处理器性能状态等级越低其消耗的能量也就越低。
传统的处理器节能方案通常是由操作系统的电源策略或中央处理器(Central Processing Unit，简称为CPU)调频策略来完成，通过控制处理器C-State和性能状态等级来达到节能。操作系统在运行时会根据当前处理器负载状况调整处理器性能状态等级以达到满足当前使用的前提下降低处理器能量。例如，Windows操作系统通过设置电源计划来配置操作系统节能，Linux操作系统可通过配置CPU调频策略来控制处理器性能状态等级达到节能，但操作系统的性能状态调节往往变化幅度过大，省电方面表现只是一般水平，在性能和节能两方面上达不到最佳，传统的节能方法在精度控制及能效利用率方面表现都较差。
针对相关技术中存在的上述问题，目前尚未提出有效的解决方案。
发明内容
本公开实施例提供了一种处理器的性能状态的调节方法、装置、UEFI、存储介质及电子装置，以至少解决相关技术中存在的对处理器的性能状态调节精度低，进而容易导致能效利用率低的问题。
根据本公开的一个实施例，提供了一种处理器的性能状态的调节方法，包括：获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；将所述处理器的性能状态值调整为所述目标性能状态值。
根据本公开的另一个实施例，提供了一种处理器的性能状态的调节装置，包括：获取模块，设置为获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；处理模块，设置为依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；调整模块，设置为将所述处理器的性能状态值调整为所述目标性能状态值。
根据本公开的另一个实施例，提供了一种统一的可扩展固件接口UEFI，包括上述实施例所述的调节装置。
根据本公开的又一个实施例，还提供了一种存储介质，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行上述任一项方法实施例中的步骤。
根据本公开的又一个实施例，还提供了一种电子装置，包括存储器和处理器，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行上述任一项方法实施例中的步骤。
通过本公开，可以根据处理器的繁忙比和利用率来确定处理器的目标性能状态值，并且可以将处理器的性能状态值调整为该目标性能状态值，进而可以实现及时地将处理器的性能状态调整为与当前的繁忙比和利用率对应的性能状态值，有效避免调节幅度过大，能效浪费的问题，达到提高处理器的性能状态的调节精度以及提高能效利用率的效果。
附图说明
此处所说明的附图用来提供对本公开的进一步理解，构成本申请的一部分，本公开的示意性实施例及其说明用于解释本公开，并不构成对本公开的不当限定。在附图中：
图1是根据本公开实施例的处理器的性能状态的调节方法的流程图；
图2是根据本公开实施例的UEFI固件控制处理器性能状态的方法流程图；
图3是根据本公开实施例的UEFI固件调节性能状态的方法的流程图；
图4是根据本公开实施例的UEFI固件实时调节处理器性能状态的方法流程图；
图5是根据本公开实施例的处理器繁忙比获取模块所执行的操作流程图；
图6是根据本公开实施例的处理器性能状态调节模块的结构示意图；
图7是根据本公开实施例处理器性能状态调节模块所执行的操作流程图；
图8是根据本公开实施例的处理器的性能状态的调节装置的结构框图。
具体实施方式
下文中将参考附图并结合实施例来详细说明本公开。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。
需要说明的是，本公开的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。
在本实施例中提供了一种处理器的性能状态的调节方法，图1是根据本公开实施例的处理器的性能状态的调节方法的流程图，如图1所示，该流程包括如下步骤：
步骤S102，获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；
步骤S104，依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；
步骤S106，将所述处理器的性能状态值调整为所述目标性能状态值。
可选地，执行上述操作的可以是独立于OS系统或系统下软件之外的统一的可扩展固件接口(Unified Extensible Firmware Interface，简称为UEFI)或类似功能模块管理。在得到所述目标性能状态值时，可以根据处理器的繁忙比，基于利用率守恒准则，同时结合当前使用的不同能效模式根据应用压力大小分析，能够达到精确化控制处理器性能状态，其中，能效模式可按照节能和性能的要求划分为不同模式。上述的繁忙比为：BUSY_x＝ΔC0/P_Current；上述的利用率守恒公式为：USAGE_x＝ΔC0/P_Max＝USAGE_x'＝ΔC0'/P_Max；其中，ΔC0为CORE资源增量数值，P_Max为处理器支持的最大性能状态值。需要说明的是，在利用率守恒的前提下，处理器在下一时刻的利用率USAGE_x'等同于当前时刻的利用率USAGE_x，基于此原理得到处理器利用率守恒公式为：USAGE_x＝ΔC0/P_Max＝USAGE_x'＝ΔC0'/P_Max。所以处理器下一时刻的Core资源增量数值和当前时刻一样，即处理器Core资源增量数值不变。基于USAGE_x利用率守恒原理，即可计算BUSY_x所需要的 P_Current，其中BUSY_x值越大说明处理器资源利用效率就越高。其中，利用率守恒的本质就是在明确的操作系统OS资源需求(利用率需求)前提下，提升处理器主频会导致UEFI计算的繁忙比降低，降低处理器主频会导致繁忙比升高。基于USAGE_x利用率守恒，即可计算BUSY_x，进而计算所需的性能状态值。
在上述实施例中，预先生成的配置文件可以是预先通过多次试验测试所得到的各繁忙比与各个目标性能状态值的调节粒度，其中，该性能状态值的调节粒度可以是正向调节粒度，也可以是负向调节粒度，该性能状态值的调节粒度也可以替换成调节算法，同样地，该调节算法可以是正向调节算法，也可以是负向调节算法。
在上述实施例中，可以根据处理器的繁忙比来确定处理器的目标性能状态值，并且可以将处理器的性能状态值调整为该目标性能状态值，进而可以实现及时地将处理器的性能状态调整为与当前的繁忙比对应的性能状态值，有效避免调节幅度过大，能效浪费的问题，达到提高处理器的性能状态的调节精度以及提高能效利用率的效果。此外，上述步骤S102-106可以周期性执行，即，在一个周期内(或称为一个轮询周期)调整一次处理器的性能状态值。
在一个可选的实施例中，获取处理器的当前繁忙比包括：由统一的可扩展固件接口UEFI获取所述处理器的所述当前繁忙比；依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值包括：由所述UEFI依据预先生成的所述配置文件确定与所述当前繁忙比对应的所述目标性能状态值；将所述处理器的性能状态值调整为所述目标性能状态值包括：由所述UEFI将所述处理器的性能状态值调整为所述目标性能状态值。在本实施例中，可以基于UEFI固件实现服务器功效管理，不依赖于操作系统和第三方软硬件工具，通过UEFI固件的方式实现服务器处理器性能状态的实时控制。在商用操作系统之外，UEFI基于周期性的轮询，实时监测处理器的CORE增量值，计算繁忙比。然后根据繁忙比，同时结合预先生成的配置文件，能够达到精确化控制处理器性能状态，避免处理器主 频资源被浪费。同时能基于UEFI固件计算的繁忙比和利用率，直接判断轮询期间，OS是否需要快速响应，从而直接设定所需的性能状态，最大程度保证业务性能要求，降低因处理器性能状态步进缓慢导致的业务延迟问题，而当业务延迟需求时，又可实现基于利用率守恒，精确获得所需的性能状态。此外，需要说明的是，由UEFI执行上述操作仅是一种可选的实施方式，在实际应用中不仅限于此。
在一个可选的实施例中，依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值包括：确定预先创建的所述处理器的性能状态等级支持范围表，其中，所述性能状态等级支持范围表中至少记录有按照预定频率间隔划分的所述处理器所支持的主频大小，所述性能状态等级支持范围表中记录的最小主频为所述处理器所支持的最小主频，所述性能状态等级支持范围表中记录的最大主频为所述处理器所支持的最大主频；依据预先生成的配置文件确定与所述当前繁忙比对应的当前性能状态值的调整粒度；在所述性能状态等级支持范围表查找与所述当前性能状态值的调整粒度对应的主频值，并将查找到的所述主频值作为所述目标性能状态值中所包括的主频值。在本实施例中，可以根据处理器类型动态创建其支持的性能状态等级范围表，比如处理器最小主频是8，最大主频是20，创建的性能状态等级支持范围表可以为：{8,9,10,11,12,13,14,15,16,17,18,19,20}。在本实施例中，预定频率可以为8，也可以为9，或表中其他数值。
在一个可选的实施例中，在将所述处理器的性能状态值调整为所述目标性能状态值之前，所述方法还包括：在确定所述繁忙比超过预定阈值时，将所述处理器的性能状态值调整为所述处理器所支持的最大性能状态值。在本实施例中，在处理器的繁忙比过高时，例如，超过99％时，通过将处理器性能状态调整至所述处理器所支持的最大等级，可以保证处理器性能不损失，最大程度响应操作系统业务。
在一个可选的实施例中，在获取处理器的繁忙比之前，所述方法还包括：确定开启了智能功效管理IEM功能；关闭所述处理器的自主调频能 力。
在一个可选的实施例中，所述处理器的性能状态值包括所述处理器的主频值以及所述处理器的电压值。可选地，该处理器的性能状态值还可以包括其他的用于指示所述处理器的性能状态的值。
下面结合几个具体实施例对本公开进行说明：
实施例1
本公开实施例提供了一种UEFI固件(对应于前述的UEFI)控制处理器性能状态的方法，实现服务器节能提效效果，流程如图2。
因为操作系统控制处理器性能状态精度较差，所以本实施例关闭操作系统控制性能状态功能，将处理器性能状态控制完全交给UEFI固件。
步骤1.UEFI固件启动过程中判断是否开启IEM功能，如果开启则继续下面步骤。
步骤2.UEFI固件通过改变处理器与操作系统电源管控能力权限来关闭操作系统控制处理器性能状态功能。
步骤3.关闭处理器的自主调频功能。
步骤4.开启周期性轮询功能。
步骤5.注册相关处理函数。
如：UEFI固件处理函数直接操作处理器寄存器来设置其性能状态。
实施例2
本公开实施例提供了一种UEFI固件调节性能状态的方法，实现服务器节能提效效果，流程如图3。
在一个轮询周期内执行如下步骤用于调节处理器性能状态。
步骤1.根据处理器类型动态创建其支持的性能状态等级范围表。
比如，处理器最小主频是8，最大主频是20，创建的性能状态等级支 持范围表为：{8,9,10,11,12,13,14,15,16,17,18,19,20}；
步骤2.计算当前处理器性能状态下的繁忙比。
将获取CORE资源增量数值除以处理器当前性能状态值作为繁忙比结果。
步骤3.判断繁忙比是否超过最大阈值，如果大于最大阈值，则执行步骤4，否则执行步骤5。
比如计算出当前处理器繁忙比达到99％(假设设定的最大阈值为99％)，直接将处理器性能状态等级调整至表中最大值。
步骤4.将处理器性能状态调整至支持范围表中的最大等级，保证处理器性能不损失，最大程度响应操作系统业务。
步骤5.通过处理器利用率守恒原理计算出所需的性能状态值，同时对不同应用压力下的繁忙比和利用率进行长期采集，并根据不同应用压力下的数据采样和训练结果，提前生成不同的Profile(配置文件)并配置到相应的能效模式中，结合Profile计算出当前繁忙比下最优对应的性能状态值。
步骤6.查表命中最优值对应的表中性能状态值，然后直接调整处理器性能状态值到表中的值。
实施例3
本公开实施例提供了一种通过UEFI固件实时调节处理器性能状态的装置，实现服务器节能提效效果，流程如图4所示：
服务器进入操作系统后处理器性能状态控制权限已经完全交给UEFI固件，所以UEFI固件必须实时轮询获取当前处理器繁忙比，然后根据繁忙比动态调整处理器性能状态。通过该装置可以降低处理器消耗的能量。
步骤1.服务器进入操作系统后，UEFI会周期性轮询处理器状态；
步骤2.判断是否需要调整性能状态，若需要调整则转至步骤3，即， 执行UEFI固件注册的处理函数，处理函数主要执行步骤3和4，否则转至步骤1。
步骤3.执行处理器繁忙比获取模块(同后续的获取模块82)，该模块所执行的操作如图5所示。
获取处理器的繁忙比用于判断处理器占用率情况，繁忙比越大说明当前处理器占用率就越高。性能状态调节模块需要依次为依据。
步骤4.执行处理器性能状态调节模块(同后续的处理模块84和调整模块86)，该模块具体包括的单元如图6所示。
首先通过繁忙比比较单元判断当前性能状态下处理器的占用率情况，业务模型分析单元基于智能算法，对各个业务压力模型下的繁忙比进行长期采集，并评估业务压力状态，如图7。通过性能状态计算单元，精确计算出处理器当前业务压力下所需的性能状态值，实现智能化的节能适配，最后通过性能状态调整单元完成一个轮询周期内的性能状态调整。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到根据上述实施例的方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本公开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(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本公开各个实施例所述的方法。
在本实施例中还提供了一种处理器的性能状态的调节装置，该装置用于实现上述实施例及优选实施方式，已经进行过说明的不再赘述。如以下所使用的，术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。
图8是根据本公开实施例的处理器的性能状态的调节装置的结构框图，如图8所示，该装置包括：
获取模块82，设置为获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；
处理模块84，设置为依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；
调整模块86，设置为将所述处理器的性能状态值调整为所述目标性能状态值。
在一个可选的实施例中，所述处理模块84可以通过如下方式确定所述目标性能状态值：确定预先创建的所述处理器的性能状态等级支持范围表，其中，所述性能状态等级支持范围表中至少记录有按照预定频率间隔划分的所述处理器所支持的主频大小，所述性能状态等级支持范围表中记录的最小主频为所述处理器所支持的最小主频，所述性能状态等级支持范围表中记录的最大主频为所述处理器所支持的最大主频；依据预先生成的配置文件确定与所述当前繁忙比对应的当前性能状态值的调整粒度；在所述性能状态等级支持范围表查找与所述当前性能状态值的调整粒度对应的主频值，并将查找到的所述主频值作为所述目标性能状态值中所包括的主频值。
在一个可选的实施例中，所述装置还设置为在将所述处理器的性能状态值调整为所述目标性能状态值之前，在确定所述繁忙比超过预定阈值时，将所述处理器的性能状态值调整为所述处理器所支持的最大性能状态值。
在一个可选的实施例中，所述装置还设置为在获取处理器的繁忙比之前，确定开启了智能功效管理IEM功能；关闭所述处理器的自主调频能力。
在一个可选的实施例中，所述处理器的性能状态值包括所述处理器的主频值以及所述处理器的电压值。
根据本公开的另一个实施例，还提供了一种统一的可扩展固件接口UEFI，包括上述任一项实施例中所述的调节装置。
需要说明的是，上述各个模块是可以通过软件或硬件来实现的，对于后者，可以通过以下方式实现，但不限于此：上述模块均位于同一处理器中；或者，上述各个模块以任意组合的形式分别位于不同的处理器中。
本公开的实施例还提供了一种存储介质，该存储介质中存储有计算机程序，其中，该计算机程序被设置为运行时执行上述任一项方法实施例中的步骤。
可选地，在本实施例中，上述存储介质可以包括但不限于：U盘、只读存储器(Read-Only Memory，简称为ROM)、随机存取存储器(Random Access Memory，简称为RAM)、移动硬盘、磁碟或者光盘等各种可以存储计算机程序的介质。
本公开的实施例还提供了一种电子装置，包括存储器和处理器，该存储器中存储有计算机程序，该处理器被设置为运行计算机程序以执行上述任一项方法实施例中的步骤。
可选地，上述电子装置还可以包括传输设备以及输入输出设备，其中，该传输设备和上述处理器连接，该输入输出设备和上述处理器连接。
可选地，本实施例中的具体示例可以参考上述实施例及可选实施方式中所描述的示例，本实施例在此不再赘述。
显然，本领域的技术人员应该明白，上述的本公开的各模块或各步骤可以用通用的计算装置来实现，它们可以集中在单个的计算装置上，或者 分布在多个计算装置所组成的网络上，可选地，它们可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，并且在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤，或者将它们分别制作成各个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本公开不限制于任何特定的硬件和软件结合。
以上所述仅为本公开的优选实施例而已，并不用于限制本公开，对于本领域的技术人员来说，本公开可以有各种更改和变化。凡在本公开的原则之内，所作的任何修改、等同替换、改进等，均应包含在本公开的保护范围之内。
工业实用性
如上所述，本发明实施例提供的一种处理器的性能状态的调节方法、装置、UEFI、存储介质及电子装置具有以下有益效果：解决了相关技术中存在的调节幅度过大，能效浪费的问题，达到提高处理器的性能状态的调节精度以及提高能效利用率的效果。

Claims (10)

  1. 一种处理器的性能状态的调节方法，包括：
    获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；
    依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；
    将所述处理器的性能状态值调整为所述目标性能状态值。
  2. 根据权利要求1所述的方法，其中，
    获取处理器的当前繁忙比包括：由统一的可扩展固件接口UEFI获取所述处理器的所述当前繁忙比；
    依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值包括：由所述UEFI依据预先生成的所述配置文件确定与所述当前繁忙比对应的所述目标性能状态值；
    将所述处理器的性能状态值调整为所述目标性能状态值包括：由所述UEFI将所述处理器的性能状态值调整为所述目标性能状态值。
  3. 根据权利要求1所述的方法，其中，依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值包括：
    确定预先创建的所述处理器的性能状态等级支持范围表，其中，所述性能状态等级支持范围表中至少记录有按照预定频率间隔划分的所述处理器所支持的主频大小，所述性能状态等级支持范围表中记录的最小主频为所述处理器所支持的最小主频，所述性能状态等级支持范围表中记录的最大主频为所述处理器所支持的最大主频；
    依据预先生成的配置文件确定与所述当前繁忙比对应的当前性能状态值的调整粒度；
    在所述性能状态等级支持范围表查找与所述当前性能状态值的调整粒度对应的主频值，并将查找到的所述主频值作为所述目标性能状态值中所包括的主频值。
  4. 根据权利要求1所述的方法，其中，在将所述处理器的性能状态值调整为所述目标性能状态值之前，所述方法还包括：
    在确定所述繁忙比超过预定阈值时，将所述处理器的性能状态值调整为所述处理器所支持的最大性能状态值。
  5. 根据权利要求1所述的方法，其中，在获取处理器的繁忙比之前，所述方法还包括：
    确定开启了智能功效管理IEM功能；
    关闭所述处理器的自主调频能力。
  6. 根据权利要求1所述的方法，其中，所述处理器的性能状态值包括所述处理器的主频值以及所述处理器的电压值。
  7. 一种处理器的性能状态的调节装置，包括：
    获取模块，设置为获取处理器的当前繁忙比，其中，所述当前繁忙比为所述处理器核CORE当前资源增量数值与所述处理器当前性能状态值的比值；
    处理模块，设置为依据预先生成的配置文件确定与所述当前繁忙比对应的目标性能状态值，其中，所述配置文件中记录有繁忙比与性能状态值的调节粒度；
    调整模块，设置为将所述处理器的性能状态值调整为所述目标性能状态值。
  8. 一种统一的可扩展固件接口UEFI，包括：权利要求7所述的调节装置。
  9. 一种存储介质，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行所述权利要求1至6任一项中所述的方法。
  10. 一种电子装置，包括存储器和处理器，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行所述权利要求1至6任一项中所述的方法。
PCT/CN2020/079022 2019-05-31 2020-03-12 处理器的性能状态的调节方法、装置、uefi及存储介质 WO2020238326A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910473230.2A CN112015260A (zh) 2019-05-31 2019-05-31 处理器的性能状态的调节方法、装置、uefi及存储介质
CN201910473230.2 2019-05-31

Publications (1)

Publication Number Publication Date
WO2020238326A1 true WO2020238326A1 (zh) 2020-12-03

Family

ID=73506399

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/079022 WO2020238326A1 (zh) 2019-05-31 2020-03-12 处理器的性能状态的调节方法、装置、uefi及存储介质

Country Status (2)

Country Link
CN (1) CN112015260A (zh)
WO (1) WO2020238326A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115374164B (zh) * 2022-10-20 2023-04-28 深圳比特微电子科技有限公司 搜频调速方法、装置和数据处理设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6076171A (en) * 1997-03-28 2000-06-13 Mitsubishi Denki Kabushiki Kaisha Information processing apparatus with CPU-load-based clock frequency
US20100037080A1 (en) * 2008-08-05 2010-02-11 Kabushiki Kaisha Toshiba Portable terminal device
CN102662822A (zh) * 2012-04-26 2012-09-12 华为技术有限公司 负载监控装置和方法
CN103716694A (zh) * 2013-12-09 2014-04-09 乐视致新电子科技（天津）有限公司 智能家电、智能家电cpu变频方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1855793A (zh) * 2005-04-19 2006-11-01 华为技术有限公司 一种视音频编码速度的控制方法
US7917789B2 (en) * 2007-09-28 2011-03-29 Intel Corporation System and method for selecting optimal processor performance levels by using processor hardware feedback mechanisms
US8909962B2 (en) * 2009-12-16 2014-12-09 Qualcomm Incorporated System and method for controlling central processing unit power with guaranteed transient deadlines
CN105740073A (zh) * 2016-01-21 2016-07-06 浪潮(北京)电子信息产业有限公司 一种动态控制操作系统进程数量的方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6076171A (en) * 1997-03-28 2000-06-13 Mitsubishi Denki Kabushiki Kaisha Information processing apparatus with CPU-load-based clock frequency
US20100037080A1 (en) * 2008-08-05 2010-02-11 Kabushiki Kaisha Toshiba Portable terminal device
CN102662822A (zh) * 2012-04-26 2012-09-12 华为技术有限公司 负载监控装置和方法
CN103716694A (zh) * 2013-12-09 2014-04-09 乐视致新电子科技（天津）有限公司 智能家电、智能家电cpu变频方法及装置

Also Published As

Publication number Publication date
CN112015260A (zh) 2020-12-01

Similar Documents

Publication Publication Date Title
US8473768B2 (en) Power control apparatus and method for cluster system
US7814485B2 (en) System and method for adaptive power management based on processor utilization and cache misses
US9176572B2 (en) System and method for controlling central processing unit power with guaranteed transient deadlines
US9026816B2 (en) Method and system for determining an energy-efficient operating point of a platform
US9760154B2 (en) Method of dynamically controlling power in multicore environment
US9026822B2 (en) Dynamically adjusting operating frequency of a arithemetic processing device for predetermined applications based on power consumption of the memory in real time
CN101853066A (zh) 一种自动实时调整系统时钟频率的方法和装置
US20160239343A1 (en) Multi-tenant monitoring
CN105518563A (zh) 动态电压频率调整系统及其方法
US7992015B2 (en) Processor performance state optimization
CN103345415A (zh) 管理定时器的终端节能方法和装置
CN105511593A (zh) 一种用于Linux系统的CPU子系统频率调节方法和装置
WO2017101362A1 (zh) 一种智能终端的功耗管控方法及系统
KR102060431B1 (ko) 멀티 코어 시스템의 전력 관리 장치 및 방법
CN103108039B (zh) 一种低能耗集群环境下的服务质量保证方法
EP3649536A1 (en) Core frequency management using effective utilization for power-efficient performance
US20140282587A1 (en) Multi-core binary translation task processing
CN113075982A (zh) 一种服务器智能网卡散热方法、装置、系统及介质
WO2020238326A1 (zh) 处理器的性能状态的调节方法、装置、uefi及存储介质
CN103092328A (zh) 一种基于磁盘休眠的计算机节能的方法
CN111443792B (zh) 处理器的功耗控制方法、装置以及电子和存储设备
CN111246549B (zh) 一种节点休眠、唤醒时间提供的方法及装置
CN114327023B (zh) 一种Kubernetes集群的节能方法、系统、计算机介质和电子设备
Eibel et al. Empya: Saving energy in the face of varying workloads
CN104536833A (zh) 一种提高高频交易性能的方法

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: 20812959

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20812959

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 20812959

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 17.05.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20812959

Country of ref document: EP

Kind code of ref document: A1