The DOI matches its Crossref record.
Rapid and accurate energy models through calibration with IPMI and RAPL
Abstract
Summary: Energy consumption in Cloud and High Performance Computing platforms is a significant issue and affects aspects such as the cost of energy and the cooling of the data center. Host level monitoring and prediction provides the groundwork for improving energy efficiency through the placement of workloads. Monitoring must be fast and efficient without unnecessary overhead, to enable scalability. This precludes the use of Watt meters attached per host, requiring alternative approaches such as integrated measurements and models. IPMI and RAPL are subject to error and partial measurement, which may be mitigated. Models allow for prediction and more responsive measures of power consumption, but require calibrating. The causes of calibration error are discussed, along with mitigation strategies, without overly complicating the underlying model. An outcome is a Watt meter emulator that provides hosts level power measurement along with estimated power consumption for a given workload, with an average error of 0.20W.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“A body of work has also looked at improving estimates of energy usage from RAPL by fitting a regression model to real energy usage patterns (Povoa et al., 2019; Kavanagh and Djemame, 2019; Ghosh et al., 2013; Song et al., 2013).”
From Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning (2020) · cites this paper as Kavanagh and Djemame, 2019
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Errors highly depend on the specific chipset and even the workload”
From Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning (2020) · cites this paper as Kavanagh and Djemame (2019)
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“First, we rely on downstream hardware APIs which themselves have errors.”
From Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning (2020) · cites this paper as Kavanagh and Djemame (2019)
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
No references could be resolved yet.
Research works printed but not linked: 41
- 10.1007/10968987_3 · DOI
- 10.1109/MM.2012.12 · DOI
- 10.1007/978-3-642-45249-9_7 · DOI
- HackenbergD IlscheT SchöneR MolkaD SchmidtM NagelWE.Power measurement techniques on standard compute nodes: a quantitative comparison. Paper presented at: 2013 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS);2013;Austin TX. · DOI
- DesrochersS ParadisC WeaverVM.A validation of DRAM RAPL power measurements. In: Proceedings of the Second International Symposium on Memory Systems (MEMSYS);2016;Alexandria VA. · DOI · Paper page
- 10.1145/2425248.2425252 · DOI
- McCrawH RalphJ DanalisA DongarraJ.Power monitoring with PAPI for extreme scale architectures and dataflow‐based programming models. Paper presented at: 2014 IEEE International Conference on Cluster Computing (CLUSTER);2014;Madrid Spain. · DOI
- WilkeC GötzS RichlyS.JouleUnit: a generic framework for software energy profiling and testing. In: Proceedings of the 2013 Workshop on Green in/by Software Engineering (GIBSE);2013;Fukuoka Japan. · DOI
- 10.1016/j.jpdc.2014.06.007 · DOI
- KansalA ZhaoF LiuJ KothariN BhattacharyaAA.Virtual machine power metering and provisioning. In: Proceedings of the 1st ACM Symposium on Cloud Computing (SoCC);2010;Indianapolis IN. · DOI
- 10.1016/j.simpat.2013.05.002 · DOI
- LarosJH PokornyP DeBonisD.PowerInsight ‐ a commodity power measurement capability. Paper presented at: 2013 International Green Computing Conference Proceedings (IGCC);2013;Arlington VA. · DOI
- 10.1109/TPDS.2009.76 · DOI
- BohraAEH ChaudharyV.VMeter: power modelling for virtualized clouds. Paper presented at: 2010 IEEE International Symposium on Parallel & Distributed Processing Workshops and Phd Forum (IPDPSW);2010;Atlanta GA.
- 10.1016/j.future.2013.07.008 · DOI
- SmithJW Khajeh‐HosseiniA WardJS SommervilleI.CloudMonitor: profiling power usage. Paper presented at: 2012 IEEE Fifth International Conference on Cloud Computing;2012;Honolulu HI. · DOI
- FarahnakianF LiljebergP PlosilaJ.LiRCUP: linear regression based CPU usage prediction algorithm for live migration of virtual machines in data centers. Paper presented at: 2013 39th Euromicro Conference on Software Engineering and Advanced Applications (SEAA);2013;Santander Spain. · DOI
- SchubertS KosticD ZwaenepoelW ShinKG.Profiling software for energy consumption. Paper presented at: 2012 IEEE International Conference on Green Computing and Communications;2012;Besançon France. · DOI
- JiangZ LuC CaiY JiangZ MaC.VPower: metering power consumption of VM. Paper presented at: 2013 IEEE 4th International Conference on Software Engineering and Service Science;2013;Beijing China.
- KhanA YanX TaoS AnerousisN.Workload characterization and prediction in the cloud: a multiple time series approach. Paper presented at: 2012 IEEE Network Operations and Management Symposium;2012;Maui HI. · DOI
- Alzamil I, 2016, Economics of Grids, Clouds, Systems, and Services: 13th International Conference, GECON 2016, Athens, Greece, September 20‐22, 2016, Revised Selected Papers
- RoyN DubeyA GokhaleA.Efficient autoscaling in the cloud using predictive models for workload forecasting. Paper presented at: 2011 IEEE 4th International Conference on Cloud Computing;2011;Washington DC. · DOI
- 10.1016/j.future.2011.05.027 · DOI
- QuirozA KimH ParasharM GnanasambandamN SharmaN.Towards autonomic workload provisioning for enterprise grids and clouds. Paper presented at: 2009 10th IEEE/ACM International Conference on Grid Computing;2009;Banff Canada. · DOI
- Djemame K, 2016, Economics of Grids, Clouds, Systems, and Services: 13th International Conference, GECON 2016, Athens, Greece, September 20‐22, 2016, Revised Selected Papers
- PrevostJJ NagothuK JamshidiM KelleyB.Optimal calculation overhead for energy efficient cloud workload prediction. Paper presented at: 2014 World Automation Congress (WAC);2014;Waikoloa Village HI. · DOI
- AdhinarayananV SubramaniamB FengW‐C.Online power estimation of graphics processing units. Paper presented at: 2016 16th IEEE/ACM International Symposium on Cluster Cloud and Grid Computing (CCGrid);2016;Cartagena Colombia. · DOI
- JuurlinkB LucasJ MammeriN et al.Enabling GPU software developers to optimize their applications—the LPGPU2 approach. Paper presented at: 2017 Conference on Design and Architectures for Signal and Image Processing (DASIP);2017;Dresden Germany. · DOI
- SongS SuC RountreeB CameronKW.A simplified and accurate model of power‐performance efficiency on emergent GPU architectures. Paper presented at: 2013 IEEE 27th International Symposium on Parallel and Distributed Processing;2013;Boston MA. · DOI
- 10.1145/2508148.2485964 · DOI
- 10.1007/s11227-016-1680-4 · DOI
- MeiX ChuX LiuH LeungYW LiZ.Energy efficient real‐time task scheduling on CPU‐GPU hybrid clusters. Paper presented at: IEEE INFOCOM 2017 ‐ IEEE Conference on Computer Communications;2017;Atlanta GA. · DOI
- RossigneuxF LefevreL GelasJP De AssuncaoMD.A generic and extensible framework for monitoring energy consumption of OpenStack clouds. Paper presented at: 2014 IEEE Fourth International Conference on Big Data and Cloud Computing;2014;Sydney Australia. · DOI
- 10.1145/3177754 · DOI · Paper page
- HackenbergD IlscheT SchuchartJ et al.HDEEM: high definition energy efficiency monitoring. Paper presented at: 2014 Energy Efficient Supercomputing Workshop;2014;New Orleans LA. · DOI
- KavanaghR ArmstrongD DjemameK.Accuracy of energy model calibration with IPMI. Paper presented at: 2016 IEEE 9th International Conference on Cloud Computing (CLOUD);2016;San Francisco CA. · DOI
- DjemameK ArmstrongD KavanaghR et al.Energy efficiency embedded service lifecycle: towards an energy efficient cloud computing architecture. In: Proceedings of the Energy Efficient Systems (EES'2014) Workshop 2nd International Conference on ICT for Sustainability;2014;Stockholm Sweden.
- 10.1007/978-3-319-43177-2_13 · DOI
- 10.1109/TSUSC.2017.2719159 · DOI
- Arjona ArocaJ ChatzipapasA Fernández AntaA MancusoV.A measurement‐based analysis of the energy consumption of data center servers. In: Proceedings of the 5th International Conference on Future Energy Systems (e‐Energy);2014;Cambridge UK. · DOI
- WanJ WangH TanSXD et al.A fast full‐chip static power estimation method. Paper presented at: 2016 13th IEEE International Conference on Solid‐State and Integrated Circuit Technology (ICSICT);2016;Hangzhou China.
Cited by
Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning
Henderson, P., Hu, J., Romoff, J., Brunskill, E., Jurafsky, D., & Pineau, J. (2020) · arXiv · doi:10.48550/arxiv.2002.05651 · Cited by 393
“34 A body of work has also looked at improving estimates of energy usage from RAPL by fitting a regression model to real energy usage patterns (Povoa et al., 2019; Kavanagh and Djemame, 2019; Ghosh et al., 2013; Song et al., 2013).”
Background
Metadata
- Authors
- Kavanagh, R., & Djemame, K.
- Affiliations
- University of Leeds (GB)
- Corresponding author
- Kavanagh, Richard
- Venue
- Concurrency and Computation: Practice and Experience
- Publisher
- Wiley
- Published
- 2019-01-09
- Volume
- 31
- Issue
- 13
- ISSN
- 1532-0626 (ISSN-L)
- 1532-0634
- OpenAlex
- W2924789150
- Cited by
- 23
- References
- 0
- Citations by year
- 2026: 1 · 2025: 5 · 2024: 3 · 2023: 4 · 2022: 5 · 2021: 2 · 2020: 2 · 2019: 1
- FWCI
- 2.54
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Concurrency and Computation Practice and Experience (publishedVersion)
- White Rose Research Online (University of Leeds, The University of Sheffield, University of York) (acceptedVersion)
- Funders
- H2020 LEIT Information and Communication Technologies (687584)
- Topics
- Physical Sciences › Computer Science › Hardware and Architecture › Energy Efficiency in Computing
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
- Keywords
- energy modeling, load forecasting, IPMI, RAPL, model calibration, calibration error, cloud computing energy efficiency, workload placement, data center cooling
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
