Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312754493> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W4312754493 abstract "Power capping is an important technique for high-density servers to safely oversubscribe the power infrastructure in a data center. However, power capping is commonly accomplished by dynamically lowering the server processors’ frequency levels, which can result in degraded application performance. For servers that run important machine learning (ML) applications with Service-Level Objective (SLO) requirements, inference performance such as recognition accuracy must be optimized within a certain latency constraint, which demands high server performance. In order to achieve the best inference accuracy under the desired latency and server power constraints, this paper proposes OptimML, a multi-input-multi-output (MIMO) control framework that jointly controls both inference latency and server power consumption, by flexibly adjusting the machine learning model size (and so its required computing resources) when server frequency needs to be lowered for power capping. Our results on a hardware testbed show that OptimML achieves higher inference accuracy compared with several well-designed baselines, while respecting both latency and power constraints." @default.
- W4312754493 created "2023-01-05" @default.
- W4312754493 creator A5028960226 @default.
- W4312754493 creator A5074281657 @default.
- W4312754493 date "2022-07-01" @default.
- W4312754493 modified "2023-10-16" @default.
- W4312754493 title "Performance Optimization of Machine Learning Inference under Latency and Server Power Constraints" @default.
- W4312754493 cites W1536680647 @default.
- W4312754493 cites W1919191429 @default.
- W4312754493 cites W1964622779 @default.
- W4312754493 cites W1969742965 @default.
- W4312754493 cites W1987912519 @default.
- W4312754493 cites W1993508117 @default.
- W4312754493 cites W2015244008 @default.
- W4312754493 cites W2033597569 @default.
- W4312754493 cites W2038337218 @default.
- W4312754493 cites W2038509324 @default.
- W4312754493 cites W2099786809 @default.
- W4312754493 cites W2100948980 @default.
- W4312754493 cites W2112806194 @default.
- W4312754493 cites W2133248872 @default.
- W4312754493 cites W2156756912 @default.
- W4312754493 cites W2172654076 @default.
- W4312754493 cites W2194775991 @default.
- W4312754493 cites W2543907405 @default.
- W4312754493 cites W2554302513 @default.
- W4312754493 cites W2772948367 @default.
- W4312754493 cites W2884675507 @default.
- W4312754493 cites W2928790192 @default.
- W4312754493 cites W2951001719 @default.
- W4312754493 cites W2963393494 @default.
- W4312754493 cites W2966297916 @default.
- W4312754493 cites W3128172237 @default.
- W4312754493 cites W3139989176 @default.
- W4312754493 doi "https://doi.org/10.1109/icdcs54860.2022.00039" @default.
- W4312754493 hasPublicationYear "2022" @default.
- W4312754493 type Work @default.
- W4312754493 citedByCount "1" @default.
- W4312754493 crossrefType "proceedings-article" @default.
- W4312754493 hasAuthorship W4312754493A5028960226 @default.
- W4312754493 hasAuthorship W4312754493A5074281657 @default.
- W4312754493 hasConcept C120314980 @default.
- W4312754493 hasConcept C121332964 @default.
- W4312754493 hasConcept C149635348 @default.
- W4312754493 hasConcept C154945302 @default.
- W4312754493 hasConcept C163258240 @default.
- W4312754493 hasConcept C168292644 @default.
- W4312754493 hasConcept C2776214188 @default.
- W4312754493 hasConcept C2984118289 @default.
- W4312754493 hasConcept C31258907 @default.
- W4312754493 hasConcept C31395832 @default.
- W4312754493 hasConcept C41008148 @default.
- W4312754493 hasConcept C62520636 @default.
- W4312754493 hasConcept C76155785 @default.
- W4312754493 hasConcept C79403827 @default.
- W4312754493 hasConcept C82876162 @default.
- W4312754493 hasConcept C93996380 @default.
- W4312754493 hasConceptScore W4312754493C120314980 @default.
- W4312754493 hasConceptScore W4312754493C121332964 @default.
- W4312754493 hasConceptScore W4312754493C149635348 @default.
- W4312754493 hasConceptScore W4312754493C154945302 @default.
- W4312754493 hasConceptScore W4312754493C163258240 @default.
- W4312754493 hasConceptScore W4312754493C168292644 @default.
- W4312754493 hasConceptScore W4312754493C2776214188 @default.
- W4312754493 hasConceptScore W4312754493C2984118289 @default.
- W4312754493 hasConceptScore W4312754493C31258907 @default.
- W4312754493 hasConceptScore W4312754493C31395832 @default.
- W4312754493 hasConceptScore W4312754493C41008148 @default.
- W4312754493 hasConceptScore W4312754493C62520636 @default.
- W4312754493 hasConceptScore W4312754493C76155785 @default.
- W4312754493 hasConceptScore W4312754493C79403827 @default.
- W4312754493 hasConceptScore W4312754493C82876162 @default.
- W4312754493 hasConceptScore W4312754493C93996380 @default.
- W4312754493 hasLocation W43127544931 @default.
- W4312754493 hasOpenAccess W4312754493 @default.
- W4312754493 hasPrimaryLocation W43127544931 @default.
- W4312754493 hasRelatedWork W2044364864 @default.
- W4312754493 hasRelatedWork W2065450024 @default.
- W4312754493 hasRelatedWork W2152313554 @default.
- W4312754493 hasRelatedWork W2159099865 @default.
- W4312754493 hasRelatedWork W2171408034 @default.
- W4312754493 hasRelatedWork W2883256816 @default.
- W4312754493 hasRelatedWork W2998256594 @default.
- W4312754493 hasRelatedWork W3003320923 @default.
- W4312754493 hasRelatedWork W3092582874 @default.
- W4312754493 hasRelatedWork W2106140982 @default.
- W4312754493 isParatext "false" @default.
- W4312754493 isRetracted "false" @default.
- W4312754493 workType "article" @default.