IP Library › Granted Patent US 12,730,674
Granted Patent B2
US 12,730,674 · App. 16/904,621 · Granted Sep 8, 2026

Power-performance based system management

Inventors: Yang Liu (Shanghai, CN); Yue Xu (Shanghai, CN); Peng Fei Gou (Shanghai, CN); Meng Li (Changshu, CN); Xing Zhao (Shanghai, CN)
Assignee: International Business Machines Corporation
G06F9/4893G06F11/3495G06F2209/486
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,730,674
App. No.
16/904,621
Granted
Sep 8, 2026
Kind
B2
Abstract

A method comprises receiving a workload for a computer system; sweeping at least one parameter of the computer system while executing the workload; monitoring one or more characteristics of the computer system while sweeping the at least one parameter, the one or more characteristics including total power consumption of the computer system; generating a power profile for the workload that indicates a respective selected value for the at least one parameter based on analysis of the monitored total power consumption of the computer system while sweeping the at least one parameter; and executing the workload based on the respective selected value of the at least one parameter.

Claims (61)

1 . A method comprising:

receiving a workload for a computer system;

sweeping a plurality of parameters of the computer system while executing the workload;

monitoring one or more characteristics of the computer system while sweeping the plurality of parameters, the one or more characteristics including total power consumption and performance measurement of the computer system;

generating a power profile for the workload that specifies a single value for each parameter of the plurality of parameters, the single value for each parameter selected to maximize performance per power usage based on analysis of the monitored total power consumption and performance measurement of the computer system while sweeping the plurality of parameters; and

executing the workload concurrently with one or more other workloads based on the specified value of each parameter of the plurality of parameters in the power profile being within a predefined threshold of corresponding specified values of parameters in power profiles of the one or more other workloads.

2 . The method of claim 1 , further comprising receiving one or more constraints on the plurality of parameters of the computer system.

3 . The method of claim 1 , further comprising dividing the workload into two or more stages;

wherein sweeping the plurality of parameters comprises sweeping the plurality of parameters for each of the two or more stages;

wherein monitoring the one or more characteristics comprises monitoring the one or more characteristics while sweeping the plurality of parameters for each of the two or more stages; and

wherein generating a power profile comprises generating a respective power profile for each of the two or more stages.

4 . The method of claim 1 , wherein sweeping the plurality of parameters comprises sweeping at least two parameters selected from the group consisting of central processing unit (CPU) frequency, graphics processing unit (GPU) frequency, and number of active cores in a multi-core processor.

5 . The method of claim 1 , wherein the monitored one or more characteristics of the computer system include one or more of central processing unit (CPU) power usage, graphics processing unit (GPU) power usage, fan power usage, memory power usage, disk power usage, memory bandwidth, memory latency, disk input/output bandwidth, and network bandwidth.

6 . The method of claim 1 , further comprising receiving an initial power profile for the workload; and

wherein generating the power profile comprises updating the initial power profile based on analysis of the monitored total power consumption of the computer system while sweeping the plurality of parameters.

7 . A computer management system comprising:

a storage device; and

a processor communicatively coupled to the storage device, wherein the processor is configured to:

receive a workload for a computer system;

iteratively adjust a plurality of parameters of the computer system while the workload is executed;

monitor one or more characteristics of the computer system while adjusting the plurality of parameters, the one or more characteristics including total power consumption and performance measurement of the computer system;

generate a power profile for the workload that specifies a single value for each parameter of the plurality of parameters, the single value for each parameter selected to maximize performance per power usage based on analysis of the monitored total power consumption and performance measurement of the computer system while sweeping the plurality of parameters;

store the power profile on the storage device; and

execute the workload concurrently with one or more other workloads based on the specified value of each parameter of the plurality of parameters in the power profile being within a predefined threshold of corresponding specified values of parameters in power profiles of the one or more other workloads.

8 . The computer management system of claim 7 , wherein the processor is further configured to receive one or more constraints on the plurality of parameters of the computer system.

9 . The computer management system of claim 7 , wherein the processor is further configured to:

divide the workload into two or more stages;

iteratively adjust the plurality of parameters for each of the two or more stages;

monitor the one or more characteristics while adjusting the plurality of parameters for each of the two or more stages; and

generate a respective power profile for each of the two or more stages.

10 . The computer management system of claim 7 , wherein the processor is configured to iteratively adjust at least two parameters selected from the group consisting of central processing unit (CPU) frequency, graphics processing unit (GPU) frequency, number of active cores in a multi-core processor.

11 . The computer management system of claim 7 , wherein the monitored one or more characteristics of the computer system include one or more of central processing unit (CPU) power usage, graphics processing unit (GPU) power usage, fan power usage, memory power usage, disk power usage, memory bandwidth, memory latency, disk input/output bandwidth, and network bandwidth.

12 . The computer management system of claim 7 , wherein the processor is further configured to:

receive an initial power profile for the workload; and

update the initial power profile based on analysis of the monitored total power consumption of the computer system while adjusting the plurality of parameters.

13 . A computer management system comprising:

a power-performance management engine configured to:

sweep a plurality of parameters of a computer system while a workload is executed;

monitor one or more characteristics of the computer system while sweeping the plurality of parameters, the one or more characteristics including total power consumption and performance measurement of the computer system; and

generate a power profile for the workload that specifies a single value for each parameter of the plurality of parameters, the single value for each parameter selected to maximize performance per power usage based on analysis of the monitored total power consumption and performance measurement of the computer system; and

a power-performance workload scheduler configured to schedule the workload for execution concurrently with one or more other workloads based on the specified value of each parameter of the plurality of parameters in the generated power profile being within a predefined threshold of corresponding specified values of parameters in power profiles of the one or more other workloads.

14 . A method comprising:

comparing, by a power-performance management engine (PPME), respective power performance tables for each of a plurality of workloads, each power performance table specifying a single value for at least one parameter of a plurality of parameters of a computer system for executing the respective workload, and wherein the specified values of the plurality of parameters are selected to maximize performance per power usage based on monitoring one or more characteristics of the computer system, including power consumption and performance measurement of the computer system, while iteratively adjusting the plurality of parameters;

identifying, by the PPME, at least two compatible workloads based on determining respective specified values for the plurality of parameters in the respective power performance tables are within a predefined threshold from the comparison of the respective power performance tables;

scheduling, by a power-performance workload scheduler, the at least two compatible workloads to be executed concurrently by the a second computer system; and

concurrently executing the at least two compatible workloads on the second computer system based on the respective specified values for the plurality of parameters in the respective power performance tables being within a predefined threshold from the comparison of the respective power performance tables.

15 . The method of claim 14 , wherein the plurality of parameters includes at least two parameters selected from the group consisting of central processing unit (CPU) frequency, graphics processing unit (GPU) frequency, number of active cores in a multi-core processor.

16 . The method of claim 14 , wherein the one or more monitored characteristics include one or more of central processing unit (CPU) power usage, graphics processing unit (GPU) power usage, fan power usage, memory power usage, disk power usage, memory bandwidth, memory latency, disk input/output bandwidth, and network bandwidth.

17 . A computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed by a processor, causes the processor to:

iteratively adjust plurality of parameters of a computer system while a workload is executed;

monitor one or more characteristics of the computer system while adjusting the plurality of parameters, the one or more characteristics including total power consumption and performance measurement of the computer system;

generate a power profile for the workload that specifies a single value for each parameter of the plurality of parameters, the single value for each parameter selected to maximize performance per power usage based on analysis of the monitored total power consumption and performance measurement of the computer system; and

execute the workload concurrently with one or more other workloads based on the specified value of each parameter of the plurality of parameters in the generated power profile being within a predefined threshold of corresponding specified values of parameters in power profiles of the one or more other workloads.

18 . The computer program product of claim 17 , wherein the one or more monitored characteristics include one or more of central processing unit (CPU) power usage, graphics processing unit (GPU) power usage, fan power usage, memory power usage, disk power usage, memory bandwidth, memory latency, disk input/output bandwidth, and network bandwidth.

19 . The computer program product of claim 17 , wherein the computer readable program is further configured to cause the processor to iteratively adjust the plurality of parameters in accordance with one or more constraints on the plurality of parameters of the computer system.

20 . The computer program product of claim 17 , wherein the computer readable program is further configured to cause the processor to:

divide the workload into two or more stages;

iteratively adjust the plurality of parameters for each of the two or more stages;

monitor the one or more characteristics while adjusting the plurality of parameters for each of the two or more stages; and

generate a respective power profile for each of the two or more stages.

21 . The computer program product of claim 17 , wherein the computer readable program is further configured to cause the processor to iteratively adjust at least two parameters selected from the group consisting of central processing unit (CPU) frequency, graphics processing unit (GPU) frequency, number of active cores in a multi-core processor.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 18, 2020
From: LIU, YANG; XU, YUE; GOU, PENG FEI; LI, MENG; ZHAO, XING
To: INTERNATIONAL BUSINESS MACHINES CORPORATION
Reel/Frame 052972/0222 →
Continuity (1)
Related Publication 20210397476A1 · Dec 23, 2021
References Cited (42)
US 8595515B1 · Weber · 2013 [cited by examiner]
US 9009500B1 · Fan · 2015 [cited by examiner]
US 9519878B2 · Sankar · 2016 [cited by examiner]
US 9886316B2 · Belady · 2018 [cited by examiner]
US 10429921B2 · Potlapally · 2019 [cited by applicant]
US 10558768B1 · Weber · 2020 [cited by examiner]
US 11137807B2 · Ahuja · 2021 [cited by examiner]
US 20090328055A1 · Bose · 2009 [cited by examiner]
US 20100031075A1 · Kapil · 2010 [cited by examiner]
US 20120017099A1 · David · 2012 [cited by examiner]
US 20140075222A1 · Jackson · 2014 [cited by examiner]
US 20150046685A1 · Park · 2015 [cited by applicant]
US 20160054779A1 · Bodas · 2016 [cited by applicant]
US 20160054780A1 · Bodas · 2016 [cited by examiner]
US 20160077571A1 · Sagar · 2016 [cited by examiner]
US 20160139964A1 · Chen · 2016 [cited by applicant]
US 20170031423A1 · Mandagere · 2017 [cited by applicant]
US 20170261949A1 · Hoffmann · 2017 [cited by examiner]
US 20180167878A1 · Lee · 2018 [cited by examiner]
US 20190018471A1 · Chen · 2019 [cited by examiner]
US 20190220073A1 · Wu · 2019 [cited by examiner]
US 20200142466A1 · Naik · 2020 [cited by applicant]
US 20200167190A1 · Bernat · 2020 [cited by examiner]
US 20200379533A1 · Suzuki · 2020 [cited by examiner]
CN 106502358A · 2017 [cited by applicant]
CN 109614216A · 2019 [cited by applicant]
CN 111143174A · 2020 [cited by applicant]
JP 2017531245A · 2017 [cited by applicant]
WO 2016028371A1 · 2016 [cited by applicant]
Arimilli et al., “IBM POWER9 processor and system features for computing in the cognitive era,” IBM Journal of Research and Development, vol. 62, No. 4/5, Paper 1, Sep. 24, 2018, 11 pages. <https://ieeexplore.ieee.org/a… [cited by applicant]
Bircher et al., “Complete System Power Estimation Using Processor Performance Events,” IEEE Transactions on Computers, Feb. 17, 2011, 18 pages. <https://ieeexplore.ieee.org/document/5714687?arnumber=5714687>. [cited by applicant]
Grot et al., “Optimizing Data-Center TCO with Scale-Out Processors,” IEEE Micro, Special Issue on Energy-Aware Computing, Sep. 4, 2012, 14 pages. <https://ieeexplore.ieee.org/document/6290314>. [cited by applicant]
Mell et al., “The NIST Definition of Cloud Computing,” Recommendations of the National Institute of Standards and Technology, U.S. Department of Commerce, Special Publication 800-145, Sep. 2011, 7 pages. [cited by applicant]
Rajamani et al., “Power-Performance Management on an IBM POWER7 Server,” Proceedings of the 16th ACM/IEEE international symposium on Low power electronics and design, Aug. 2010, pp. 201-206. <https://dl.acm.org/doi/abs/… [cited by applicant]
International Search Report and Written Opinion dated Sep. 23, 2021 from International Application No. PCT/IB2021/055315 filed Jun. 16, 2021. [cited by applicant]
Japan Patent Office, “Notice of Reasons for Refusal,” Dec. 17, 2024, 6 Pages, JP Application No. 2022-570218. [cited by applicant]
Intellectual Property Office, Patents Act 1977: Examination Report under Section 18(3), Jun. 11, 2025, 05 Pages, GB Application No. 2300445.0. [cited by applicant]
Response to the examination report dated Aug. 15, 2025, mailed on Aug. 29, 2025, Application No. 10-2022-7040980, 2 pages. [cited by applicant]
Intellectual Property Office, “Patents Act 1977: Search Report under Section 17(5)”, Feb. 10, 2026, 3 Pages, GB Application No. 2300445.0. [cited by applicant]
Innovation, Science and Economic Development Canada, “Office Action,” Jan. 20, 2026, 9 Pages, CA Application No. 3176180. [cited by applicant]
Intellectual Property Office, Patents Act 1977: Examination Report under Section 18(3), Aug. 15, 2025, 05 Pages, GB Application No. 2300445.0. [cited by applicant]
Intellectual Property Office, “Request for the Submission of an Opinion,” Aug. 28, 2025, 07 Pages, KR Application No. 10-2022-7040980. [cited by applicant]