Energy-minimizing workload splitting and frequency selection for guaranteed performance over heterogeneous cores
Authors: 
Aditya Priya, Rajiv Choudhury, Sujay Patni, Himkant Sharma, Moonmoon Mohanty, Krishnasuri Narayanam, Umamaheswari Devi, Pratibha Moogi, Preetam Patil, Parimal ParagAuthors Info & Claims
Aditya Priya, Rajiv Choudhury, Sujay Patni, Himkant Sharma, Moonmoon Mohanty, Krishnasuri Narayanam, Umamaheswari Devi, Pratibha Moogi, Preetam Patil, Parimal ParagAuthors Info & ClaimsPages 308 - 322
Abstract
Heterogeneous computing involves CPU architectures that support more than one core type, and it aims to achieve energy efficiency while meeting the performance guarantees. This aim can be achieved by the operating system or the on-chip driver by exploiting the differential power-performance trade-off that heterogeneous cores offer. We characterize the power-performance trade-off for an Intel CPU with heterogeneous cores and provide a mathematical framework to study heterogeneous computing. In particular, we provide probabilistic workload split and operating frequency for all active cores that allow workload execution with minimal carbon emissions. We support the analytical findings with experimental evaluations for a few representative workloads. As compared to the default Linux frequency governors, our scheme can reduce the energy-delay product by up to 80%.
References
[1]
2013. Power Management with big.LITTLE: A technical overview. https://community.arm.com/arm-community-blogs/b/architectures-and-processors-blog/posts/power-management-with-big-little-a-technical-overview.
[2]
2018. 12th Gen Intel Core Processors. https://www.intel.com/content/www/us/en/products/docs/processors/core/12th-gen-processors.html.
[3]
2018. Energy Aware Scheduling. https://docs.kernel.org/scheduler/sched-energy.html.
[4]
2022. "Performance Measurement | TensprFlow Lite". https://www.tensorflow.org/lite/performance/measurement
[5]
2023. 13th Generation Intel Core Processors Datasheet, Volume 1 of 2. https://www.intel.com/content/www/us/en/content-details/743844/13th-generation-intel-core-processors-datasheet-volume-1-of-2.html
[6]
2024. CFS Scheduler. https://docs.kernel.org/scheduler/sched-design-CFS.html.
[7]
2024. CPU Performance Scaling. https://docs.kernel.org/admin-guide/pm/cpufreq.html.
[8]
International Energy Agency. [n. d.]. Data and Statistics. Web. https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks accessed: Sep 2023.
[9]
Soren Asmussen. 1987. Applied probability and queues. Springer.
[10]
Sayed A. Banawan and Nidal M. Zeidat. 1992. A Comparative Study of Load Sharing in Heterogeneous Multicomputer Systems. In Annual Symposium on Simulation (Orlando, Florida, USA) (ANSS ’92). IEEE Computer Society Press, Washington, DC, USA, 22–31.
[11]
Luiz André Barroso and Urs Hölzle. 2007. The Case for Energy-Proportional Computing. Computer 40, 12 (2007), 33–37.
[12]
Robert Basmadjian and Hermann de Meer. 2012. Evaluating and modeling power consumption of multi-core processors. In International Conference on Future Systems (e-Energy). 1–10.
[13]
Stephen Boyd and Lieven Vandenberghe. 2004. Convex optimization. Cambridge university press.
[14]
Andre R Brodtkorb, Christopher Dyken, Trond R Hagen, Jon M Hjelmervik, and Olaf O Storaasli. 2010. State-of-the-art in heterogeneous computing. Scientific Programming 18, 1 (2010), 1–33.
[15]
Yuan-Chieh Chow and Kohler. 1979. Models for Dynamic Load Balancing in a Heterogeneous Multiple Processor System. IEEE Trans. Comput. C-28, 5 (1979), 354–361.
[16]
Standard Performance Evaluation Corporation. [n. d.]. ImageMagick: Convert, Edit, or Compose Digital Images. https://imagemagick.org/.
[17]
Standard Performance Evaluation Corporation. 2017. The SPEC CPU 2017 benchmark package. https://www.spec.org/cpu2017/.
[18]
Anthony Danalis, Gabriel Marin, Collin McCurdy, Jeremy S Meredith, Philip C Roth, Kyle Spafford, Vinod Tipparaju, and Jeffrey S Vetter. 2010. The scalable heterogeneous computing (SHOC) benchmark suite. In Workshop on general-purpose computation on graphics processing units. 63–74.
[19]
Miyuru Dayarathna, Yonggang Wen, and Rui Fan. 2016. Data Center Energy Consumption Modeling: A Survey. IEEE Communications Surveys & Tutorials 18, 1 (2016), 732–794.
[20]
Dong Du, Qingyuan Liu, Xueqiang Jiang, Yubin Xia, Binyu Zang, and Haibo Chen. 2022. Serverless Computing on Heterogeneous Computers. In International Conference on Architectural Support for Programming Languages and Operating Systems(ASPLOS ’22). ACM, 797–813.
[21]
EnerData. 2023. Between 10 and 20% of electricity consumption from the ICT sector in 2030?Web. https://www.enerdata.net/publications/executive-briefing/world-energy-consumption-from-digitalization.pdf accessed: Sep 2023.
[22]
Chris Gianos. 2023. Architecting for Flexibility and Value with Next Gen Intel® Xeon® Processors. https://hc2023.hotchips.org/assets/program/conference/day1/Platforms/HC2023.Intel.Gianos.v7.pdf
[23]
Joe Goldberg. 2018. Modern Batch Processing: A Thing of the Past or Essential Discipline?Web page. https://www.bmc.com/blogs/modern-batch-processing-thing-past-essential-discipline/ accessed: Feb 2024.
[24]
Alexander Golovin, Robert Basmadjian, Sergey Astafiev, and Alexander Rumyantsev. 2023. Little’s Law in a Single-Server System with Inactive State for Demand-Response in Data Centers with Green SLAs. In Companion Proceedings of ACM e-Energy (Orlando, FL, USA) (e-Energy ’23 Companion). Association for Computing Machinery, New York, NY, USA, 91–97.
[25]
R. Gonzalez and M. Horowitz. 1996. Energy dissipation in general purpose microprocessors. IEEE Journal of Solid-State Circuits 31, 9 (1996), 1277–1284.
[26]
Wessam M Hassanein, Layali K Rashid, and Moustafa A Hammad. 2008. Analyzing the Effects of Hyperthreading on the Performance of Data Management Systems. International Journal of Parallel Programming 36, 2 (2008), 206–225.
[27]
Mark Horowitz, Elad Alon, Dinesh Patil, Samuel Naffziger, Rajesh Kumar, and Kerry Bernstein. 2005. Scaling, power, and the future of CMOS. In IEEE InternationalElectron Devices Meeting, 2005. IEDM Technical Digest.7–15.
[28]
Bill Jones. 2023. Will Hyper-Threading Improve Processing Performance?Web page. https://www.dasher.com/will-hyper-threading-improve-processing-performance/ accessed: Feb 2024.
[29]
Aman Kansal, Feng Zhao, Jie Liu, Nupur Kothari, and Arka A. Bhattacharya. 2010. Virtual machine power metering and provisioning. In ACM Symposium on Cloud Computing(SoCC ’10). 39–50.
[30]
Kashif Nizam Khan, Mikael Hirki, Tapio Niemi, Jukka K. Nurminen, and Zhonghong Ou. 2018. RAPL in Action: Experiences in Using RAPL for Power Measurements. ACM Transactions on Modeling and Performance Evaluation of Computing Systems 3, 2 (Mar 2018), 1–26.
[31]
Donald Kinghorn. [n. d.]. Hyper-Threading may be Killing your Parallel Performance. Web page. https://www.pugetsystems.com/labs/hpc/hyper-threading-may-be-killing-your-parallel-performance-578/ accessed: Feb 2024.
[32]
R. Kumar, K.I. Farkas, N.P. Jouppi, P. Ranganathan, and D.M. Tullsen. 2003. Single-ISA heterogeneous multi-core architectures: the potential for processor power reduction. In IEEE/ACM International Symposium on Microarchitecture (MICRO-36). 81–92.
[33]
Rakesh Kumar, Dean M Tullsen, Parthasarathy Ranganathan, Norman P Jouppi, and Keith I Farkas. 2004. Single-ISA heterogeneous multi-core architectures for multithreaded workload performance. ACM SIGARCH Computer Architecture News 32, 2 (2004), 64.
[34]
Vincent Lannurien, Laurent D’Orazio, Olivier Barais, Esther Bernard, Olivier Weppe, Laurent Beaulieu, Amine Kacete, Stéphane Paquelet, and Jalil Boukhobza. 2023. HeROfake: Heterogeneous Resources Orchestration in a Serverless Cloud – An Application to Deepfake Detection. In 2023 IEEE/ACM 23rd International Symposium on Cluster, Cloud and Internet Computing (CCGrid). 154–165.
[35]
Ronald L. Larsen and Ashok K. Agrawala. 1983. Control of a Heterogeneous Two-Server Exponential Queueing System. IEEE Transactions on Software Engineering SE-9 (1983), 522–526. https://api.semanticscholar.org/CorpusID:13941700
[36]
T. Li, S. Ying, Y. Zhao, and J. Shang. 2024. Batch Jobs Load Balancing Scheduling in Cloud Computing Using Distributional Reinforcement Learning. IEEE Transactions on Parallel & Distributed Systems 35, 01 (jan 2024), 169–185.
[37]
Woei Lin and P. Kumar. 1984. Optimal control of a queueing system with two heterogeneous servers. IEEE Trans. Automat. Control 29, 8 (1984), 696–703.
[38]
Ravi Reddy Manumachu and Alexey Lastovetsky. 2022. On Energy Nonproportionality of CPUs and GPUs. In 2022 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW). 34–44.
[39]
David Meisner and Thomas F. Wenisch. 2012. DreamWeaver: Architectural Support for Deep Sleep. In International Conference on Architectural Support for Programming Languages and Operating Systems (London, England, UK) (ASPLOS XVII). Association for Computing Machinery, New York, NY, USA, 313–324.
[40]
Andreas Merkel, Jan Stoess, and Frank Bellosa. 2010. Resource-conscious scheduling for energy efficiency on multicore processors. In European Conference on Computer Systems (Paris, France) (EuroSys ’10). Association for Computing Machinery, New York, NY, USA, 153–166.
[41]
Sparsh Mittal and Jeffrey S Vetter. 2015. A survey of CPU-GPU heterogeneous computing techniques. ACM Computing Surveys (CSUR) 47, 4 (2015), 1–35.
[42]
Lucía Pons, Josué Feliu, José Puche, Chaoyi Huang, Salvador Petit, Julio Pons, María E. Gómez, and Julio Sahuquillo. 2022. Effect of Hyper-Threading in Latency-Critical Multithreaded Cloud Applications and Utilization Analysis of the Major System Resources. Future Generation Computer Systems 131 (2022), 194–208.
[43]
Thomas Rauber, Gudula Rünger, Michael Schwind, Haibin Xu, and Simon Melzner. 2014. Energy measurement, modeling, and prediction for processors with frequency scaling. The Journal of Supercomputing 70 (12 2014), 1451–1476.
[44]
Nikzad Babaii Rizvandi, Javid Taheri, and Albert Y. Zomaya. 2011. Some observations on optimal frequency selection in DVFS-based energy consumption minimization. J. Parallel and Distrib. Comput. 71, 8 (2011), 1154–1164.
[45]
Juan Carlos Saez, Alexandra Fedorova, David Koufaty, and Manuel Prieto. 2012. Leveraging core specialization via OS scheduling to improve performance on asymmetric multicore systems. ACM Transactions on Computer Systems (TOCS) 30, 2 (2012), 1–38.
[46]
Subhash Saini, Haoqiang Jin, Robert Hood, David Barker, Piyush Mehrotra, and Rupak Biswas. 2011. The impact of hyper-threading on processor resource utilization in production applications. In International Conference on High Performance Computing. 1–10.
[47]
Abul Sarwar. 1997. CMOS power consumption and Cpd calculation.Web page. https://www.ti.com/lit/an/scaa035b/scaa035b.pdf?ts=1714742752758
[48]
Claudio Scordino, Luca Abeni, and Juri Lelli. 2018. Energy-aware real-time scheduling in the linux kernel. In ACM Symposium on Applied Computing. 601–608.
[49]
Claudio Scordino, Luca Abeni, and Juri Lelli. 2019. Real-time and energy efficiency in Linux: theory and practice. ACM SIGAPP Applied Computing Review 18, 4 (2019), 18–30.
[50]
S. Shenker and A. Weinrib. 1989. The optimal control of heterogeneous queueing systems: a paradigm for load-sharing and routing. IEEE Trans. Comput. 38, 12 (1989), 1724–1735.
[51]
Jayanth Srinivasan, Sarita V. Adve, Pradip Bose, and Jude A. Rivers. 2004. The impact of technology scaling on lifetime reliability. In International Conference on Dependable Systems and Networks. 177–186.
[52]
Xueyan Tang and Samuel T Chanson. 2000. Optimizing static job scheduling in a network of heterogeneous computers. In International Conference on Parallel Processing. IEEE, 373–382.
[53]
Inc. UEFI Forum. 2022. Advanced Configuration and Power Interface ""(ACPI)"" Specification Release 6.5. https://uefi.org/sites/default/files/resources/ACPI_Spec_6_5_Aug29.pdf
[54]
Abel Weinrib and Scott Shenker. 1988. Greed is not enough: adaptive load sharing in large heterogeneous systems. IEEE INFOCOM (1988), 986–994.
[55]
Mark Whitney. 2023. Best Practices for Running HPC Batch Jobs. Web page. https://rescale.com/blog/batch-job/ accessed: Feb 2024.
[56]
Daniel Wong and Murali Annavaram. 2012. KnightShift: Scaling the Energy Proportionality Wall through Server-Level Heterogeneity. In IEEE/ACM International Symposium on Microarchitecture. 119–130.
Index Terms
- Energy-minimizing workload splitting and frequency selection for guaranteed performance over heterogeneous cores
Recommendations
Improving energy efficiency of multi-threaded applications using heterogeneous CMOS-TFET multicores
ISLPED '11: Proceedings of the 17th IEEE/ACM international symposium on Low-power electronics and designEnergy-Delay-Product-aware DVFS is a widely-used technique that improves energy efficiency by dynamically adjusting the frequencies of cores. Further, for multithreaded applications, barrier-aware DVFS is a method that can dynamically tune the ...
Parallelism via Multithreaded and Multicore CPUs
Multicore and multithreaded CPUs have become the new approach to obtaining increases in CPU performance. Numeric applications mostly benefit from a large number of computationally powerful cores. Servers typically benefit more if chip circuitry is used ...
A Novel Algorithm for Bi-objective Performance-Energy Optimization of Applications with Continuous Performance and Linear Energy Profiles on Heterogeneous HPC Platforms
Euro-Par 2021: Parallel Processing WorkshopsAbstractPerformance and energy are the two most important objectives for optimization on heterogeneous HPC platforms. This work studies a mathematical problem motivated by the bi-objective optimization of a matrix multiplication application on such ...
Comments
Information & Contributors
Information
Published In
June 2024
704 pages
ISBN:9798400704802
DOI:10.1145/3632775
Copyright © 2024 Owner/Author.
This work is licensed under a Creative Commons Attribution International 4.0 License.
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 31 May 2024
Check for updates
Author Tags
Qualifiers
- Research-article
- Research
- Refereed limited
Funding Sources
- IBM-IISc Hybrid Cloud Lab (IIHCL) open research collaboration
- Qualcomm UR 6G India
- Centre for Networked Intelligence (a Cisco Corporate Social Responsibility (CSR) Initiative)
- Science and Engineering Research Board (SERB)
- UK-India Education and Research Initiative (UKIERI)
Conference
e-Energy '24
e-Energy '24: The 15th ACM International Conference on Future and Sustainable Energy Systems
June 4 - 7, 2024
Singapore, Singapore
Acceptance Rates
Overall Acceptance Rate 160 of 446 submissions, 36%
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 266Total Downloads
- Downloads (Last 12 months)266
- Downloads (Last 6 weeks)59
Reflects downloads up to 01 Nov 2024
Other Metrics
Citations
View Options
View options
View or Download as a PDF file.
PDFeReader
View online with eReader.
eReaderHTML Format
View this article in HTML Format.
HTML FormatGet Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in