Show simple item record

dc.contributor.authorSaravanan, Karthikeyan P.
dc.contributor.authorCarpenter, Paul Matthew
dc.contributor.authorRamírez Bellido, Alejandro
dc.date.accessioned2015-05-18T11:20:14Z
dc.date.available2017-05-02T00:30:44Z
dc.date.created2014
dc.date.issued2014
dc.identifier.citationSaravanan, K.; Carpenter, P.; Alex Ramirez. A performance perspective on energy efficient HPC links. A: International Conference on Supercomputing. "ICS '14: proceedings of the 28th ACM International conference on Supercomputing". Munich: Association for Computing Machinery (ACM), 2014, p. 313-322.
dc.identifier.isbn978-1-4503-2642-1
dc.identifier.urihttp://hdl.handle.net/2117/27946
dc.description.abstractEnergy costs are an increasing part of the total cost of ownership of HPC systems. As HPC systems become increasingly energy proportional in an effort to reduce energy costs, interconnect links stand out for their inefficiency. Commodity interconnect links remain 'always-on', consuming full power even when no data is being transmitted. Although various techniques have been proposed towards energy-proportional interconnects, they are often too conservative or are not focused toward HPC. Aggressive techniques for interconnect energy savings are often not applied to HPC, in particular, because they may incur excessive performance overheads. Any energy-saving technique will only be adopted in HPC if there is no significant impact on performance, which is still the primary design objective. This paper explores interconnect energy proportionality from a performance perspective. We characterize HPC applications over on/off links and propose PerfBound, a technique that reduces link energy, subject to a bound on the application's performance degradation. We also propose PerfBoundRatio, which maintains the same performance bound across an entire hierarchical network. Finally, we propose PerfBoundPredict, which improves energy savings using an idle time prediction mechanism. Even when predictions are inaccurate, the performance degradation is still bounded. The techniques require no changes to the application and add no communication between nodes and/or switches. We evaluate our techniques using HPC traces from production supercomputers. Our results show that, configured with a 1% performance bound, 13 out of 15 applications are inside the bound, and average link energy savings are 60% for PerfBound and 68% for PerfBoundPredict.
dc.format.extent10 p.
dc.language.isoeng
dc.publisherAssociation for Computing Machinery (ACM)
dc.rightsAttribution-NonCommercial-NoDerivs 3.0 Spain
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/3.0/es/
dc.subjectÀrees temàtiques de la UPC::Informàtica::Arquitectura de computadors
dc.subjectÀrees temàtiques de la UPC::Informàtica
dc.subject.lcshHPC (Computer science)
dc.subject.lcshHigh performance computing
dc.subject.otherEnergy efficient ethernet
dc.subject.otherEnergy proportional interconnects
dc.subject.otherIdle time prediction
dc.subject.otherPerformance overhead bounding
dc.titleA performance perspective on energy efficient HPC links
dc.typeConference lecture
dc.subject.lemacInformàtica -- Estalvi d'energia
dc.contributor.groupUniversitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
dc.identifier.doi10.1145/2597652.2597671
dc.relation.publisherversionhttp://dl.acm.org/citation.cfm?id=2597671
dc.rights.accessOpen Access
local.identifier.drac15177473
dc.description.versionPostprint (published version)
local.citation.authorSaravanan, K.; Carpenter, P.; Alex Ramirez
local.citation.contributorInternational Conference on Supercomputing
local.citation.pubplaceMunich
local.citation.publicationNameICS '14: proceedings of the 28th ACM International conference on Supercomputing
local.citation.startingPage313
local.citation.endingPage322


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record