Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312998584> ?p ?o ?g. }
- W4312998584 endingPage "1328" @default.
- W4312998584 startingPage "1314" @default.
- W4312998584 abstract "Deep Neural Network (DNN) INFerence-as-a-Service (INFaaS) is the dominating workload in current data centers, for which FPGAs become promising hardware platforms because of their high flexibility and energy efficiency. The dynamic and multi-tenancy nature of INFaaS requires careful design in three aspects: multi-tenant architecture, multi-DNN scheduling, and multi-core mapping. These three factors are critical to the system latency and energy efficiency but are also challenging to optimize since they are tightly coupled and correlated. This paper proposes <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>H3M</b> , an automatic Design Space Exploration (DSE) framework to jointly optimize the <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>architecture</i> , <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>scheduling</i> , and <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>mapping</i> for serving INFaaS on cloud FPGAs. H3M explores: (1) the architecture design space with <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>H</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>eterogeneous</i> spatial <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>M</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>ulti-tenant</i> sub-accelerators, (2) layer-wise scheduling for <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>H</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>eterogeneous</i> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>M</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>ulti-DNN</i> workloads, and (3) single-layer mapping to the <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>H</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>omogeneous</i> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink/> <bold xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>M</b> <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>ulti-core</i> architecture. H3M beats state-of-the-art multi-tenant DNN accelerators, Planaria and Herald, by up to 7.5× and 3.6× in Energy-Delay-Product (EDP) reduction on the ASIC platform. On the Xilinx U200 and U280 FPGA platforms, H3M offers 2.1-5.7× and 1.8-9.0× EDP reduction over Herald." @default.
- W4312998584 created "2023-01-05" @default.
- W4312998584 creator A5015946486 @default.
- W4312998584 creator A5026831784 @default.
- W4312998584 creator A5039179408 @default.
- W4312998584 creator A5059222136 @default.
- W4312998584 creator A5059293268 @default.
- W4312998584 creator A5066747947 @default.
- W4312998584 creator A5068626165 @default.
- W4312998584 creator A5077505315 @default.
- W4312998584 date "2023-05-01" @default.
- W4312998584 modified "2023-10-14" @default.
- W4312998584 title "Serving Multi-DNN Workloads on FPGAs: A Coordinated Architecture, Scheduling, and Mapping Perspective" @default.
- W4312998584 cites W2067523571 @default.
- W4312998584 cites W2094756095 @default.
- W4312998584 cites W2183341477 @default.
- W4312998584 cites W2194775991 @default.
- W4312998584 cites W2293634267 @default.
- W4312998584 cites W2442974303 @default.
- W4312998584 cites W2600117321 @default.
- W4312998584 cites W2605347906 @default.
- W4312998584 cites W2605350416 @default.
- W4312998584 cites W2616014673 @default.
- W4312998584 cites W2883929540 @default.
- W4312998584 cites W2899915146 @default.
- W4312998584 cites W2917087921 @default.
- W4312998584 cites W2931743911 @default.
- W4312998584 cites W2935331687 @default.
- W4312998584 cites W2963163009 @default.
- W4312998584 cites W3000062206 @default.
- W4312998584 cites W3011348040 @default.
- W4312998584 cites W3016939927 @default.
- W4312998584 cites W3017521908 @default.
- W4312998584 cites W3035328498 @default.
- W4312998584 cites W3043406639 @default.
- W4312998584 cites W3043433718 @default.
- W4312998584 cites W3043571714 @default.
- W4312998584 cites W3061070527 @default.
- W4312998584 cites W3092347650 @default.
- W4312998584 cites W3101026687 @default.
- W4312998584 cites W3111640822 @default.
- W4312998584 cites W3158233068 @default.
- W4312998584 cites W3187788856 @default.
- W4312998584 cites W3213229701 @default.
- W4312998584 cites W4256333068 @default.
- W4312998584 doi "https://doi.org/10.1109/tc.2022.3214113" @default.
- W4312998584 hasPublicationYear "2023" @default.
- W4312998584 type Work @default.
- W4312998584 citedByCount "0" @default.
- W4312998584 crossrefType "journal-article" @default.
- W4312998584 hasAuthorship W4312998584A5015946486 @default.
- W4312998584 hasAuthorship W4312998584A5026831784 @default.
- W4312998584 hasAuthorship W4312998584A5039179408 @default.
- W4312998584 hasAuthorship W4312998584A5059222136 @default.
- W4312998584 hasAuthorship W4312998584A5059293268 @default.
- W4312998584 hasAuthorship W4312998584A5066747947 @default.
- W4312998584 hasAuthorship W4312998584A5068626165 @default.
- W4312998584 hasAuthorship W4312998584A5077505315 @default.
- W4312998584 hasBestOaLocation W43129985841 @default.
- W4312998584 hasConcept C123657996 @default.
- W4312998584 hasConcept C126255220 @default.
- W4312998584 hasConcept C142362112 @default.
- W4312998584 hasConcept C149635348 @default.
- W4312998584 hasConcept C153349607 @default.
- W4312998584 hasConcept C154945302 @default.
- W4312998584 hasConcept C206729178 @default.
- W4312998584 hasConcept C33923547 @default.
- W4312998584 hasConcept C41008148 @default.
- W4312998584 hasConcept C42935608 @default.
- W4312998584 hasConcept C76155785 @default.
- W4312998584 hasConcept C82876162 @default.
- W4312998584 hasConceptScore W4312998584C123657996 @default.
- W4312998584 hasConceptScore W4312998584C126255220 @default.
- W4312998584 hasConceptScore W4312998584C142362112 @default.
- W4312998584 hasConceptScore W4312998584C149635348 @default.
- W4312998584 hasConceptScore W4312998584C153349607 @default.
- W4312998584 hasConceptScore W4312998584C154945302 @default.
- W4312998584 hasConceptScore W4312998584C206729178 @default.
- W4312998584 hasConceptScore W4312998584C33923547 @default.
- W4312998584 hasConceptScore W4312998584C41008148 @default.
- W4312998584 hasConceptScore W4312998584C42935608 @default.
- W4312998584 hasConceptScore W4312998584C76155785 @default.
- W4312998584 hasConceptScore W4312998584C82876162 @default.
- W4312998584 hasFunder F4320321001 @default.
- W4312998584 hasFunder F4320329777 @default.
- W4312998584 hasIssue "5" @default.
- W4312998584 hasLocation W43129985841 @default.
- W4312998584 hasOpenAccess W4312998584 @default.
- W4312998584 hasPrimaryLocation W43129985841 @default.
- W4312998584 hasRelatedWork W1987753576 @default.
- W4312998584 hasRelatedWork W2016389538 @default.
- W4312998584 hasRelatedWork W2063534976 @default.
- W4312998584 hasRelatedWork W2330761325 @default.
- W4312998584 hasRelatedWork W2352296208 @default.
- W4312998584 hasRelatedWork W2388618054 @default.
- W4312998584 hasRelatedWork W2900316824 @default.
- W4312998584 hasRelatedWork W2995926156 @default.
- W4312998584 hasRelatedWork W3033499831 @default.