Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311463196> ?p ?o ?g. }
- W4311463196 abstract "Dynamic graph neural network (DGNN) is becoming increasingly popular because of its widespread use in capturing the dynamic features in the real world. A variety of dynamic graph neural networks designed from algorithmic perspectives have succeeded in incorporating temporal information into graph processing. Despite the promising algorithmic performance, deploying DGNNs on hardware presents additional challenges due to the model complexity, diversity, and the nature of the time-dependency. Meanwhile, the differences between DGNNs and static graph neural networks make hardware-related optimizations for static graph neural networks unsuitable for DGNNs. In this paper, We select eight prevailing DGNNs with different characteristics and profile them on both CPU and GPU. The profiling results are summarized and analyzed, providing in-depth insights into the bottlenecks of DGNNs on hardware and identifying potential optimization opportunities for future DGNN acceleration. Followed by a comprehensive survey, we provide a detailed analysis of DGNN performance bottlenecks on hardware, including temporal data dependency, workload imbalance, data movement, and GPU warm-up. We suggest several optimizations from both software and hardware perspectives. This paper is the first to provide an in-depth analysis of the hardware performance of DGNN* Code is available at https://github.com/sharc-lab/DGNN_analysis." @default.
- W4311463196 created "2022-12-26" @default.
- W4311463196 creator A5009717985 @default.
- W4311463196 creator A5030535555 @default.
- W4311463196 creator A5038753484 @default.
- W4311463196 creator A5048391735 @default.
- W4311463196 creator A5053022655 @default.
- W4311463196 date "2022-11-01" @default.
- W4311463196 modified "2023-09-27" @default.
- W4311463196 title "Bottleneck Analysis of Dynamic Graph Neural Network Inference on CPU and GPU" @default.
- W4311463196 cites W2004110412 @default.
- W4311463196 cites W2091432990 @default.
- W4311463196 cites W2136848157 @default.
- W4311463196 cites W2141125852 @default.
- W4311463196 cites W2160815625 @default.
- W4311463196 cites W2767891136 @default.
- W4311463196 cites W2794284562 @default.
- W4311463196 cites W2884001105 @default.
- W4311463196 cites W2912581782 @default.
- W4311463196 cites W2964199361 @default.
- W4311463196 cites W2997035976 @default.
- W4311463196 cites W2998313947 @default.
- W4311463196 cites W3017228913 @default.
- W4311463196 cites W3019166713 @default.
- W4311463196 cites W3024560045 @default.
- W4311463196 cites W3034795332 @default.
- W4311463196 cites W3047011367 @default.
- W4311463196 cites W3101073376 @default.
- W4311463196 cites W3101588560 @default.
- W4311463196 cites W3105753905 @default.
- W4311463196 cites W3106125969 @default.
- W4311463196 cites W3132455321 @default.
- W4311463196 cites W3134757439 @default.
- W4311463196 cites W3152893301 @default.
- W4311463196 cites W3159754263 @default.
- W4311463196 cites W3159894882 @default.
- W4311463196 cites W3162616539 @default.
- W4311463196 cites W3172512547 @default.
- W4311463196 cites W3206743063 @default.
- W4311463196 cites W4210257598 @default.
- W4311463196 doi "https://doi.org/10.1109/iiswc55918.2022.00021" @default.
- W4311463196 hasPublicationYear "2022" @default.
- W4311463196 type Work @default.
- W4311463196 citedByCount "1" @default.
- W4311463196 countsByYear W43114631962023 @default.
- W4311463196 crossrefType "proceedings-article" @default.
- W4311463196 hasAuthorship W4311463196A5009717985 @default.
- W4311463196 hasAuthorship W4311463196A5030535555 @default.
- W4311463196 hasAuthorship W4311463196A5038753484 @default.
- W4311463196 hasAuthorship W4311463196A5048391735 @default.
- W4311463196 hasAuthorship W4311463196A5053022655 @default.
- W4311463196 hasBestOaLocation W43114631962 @default.
- W4311463196 hasConcept C111919701 @default.
- W4311463196 hasConcept C113775141 @default.
- W4311463196 hasConcept C118524514 @default.
- W4311463196 hasConcept C120314980 @default.
- W4311463196 hasConcept C132525143 @default.
- W4311463196 hasConcept C149635348 @default.
- W4311463196 hasConcept C154945302 @default.
- W4311463196 hasConcept C173608175 @default.
- W4311463196 hasConcept C187191949 @default.
- W4311463196 hasConcept C199360897 @default.
- W4311463196 hasConcept C2776214188 @default.
- W4311463196 hasConcept C2777904410 @default.
- W4311463196 hasConcept C2778476105 @default.
- W4311463196 hasConcept C2780513914 @default.
- W4311463196 hasConcept C41008148 @default.
- W4311463196 hasConcept C50644808 @default.
- W4311463196 hasConcept C78766204 @default.
- W4311463196 hasConcept C80444323 @default.
- W4311463196 hasConceptScore W4311463196C111919701 @default.
- W4311463196 hasConceptScore W4311463196C113775141 @default.
- W4311463196 hasConceptScore W4311463196C118524514 @default.
- W4311463196 hasConceptScore W4311463196C120314980 @default.
- W4311463196 hasConceptScore W4311463196C132525143 @default.
- W4311463196 hasConceptScore W4311463196C149635348 @default.
- W4311463196 hasConceptScore W4311463196C154945302 @default.
- W4311463196 hasConceptScore W4311463196C173608175 @default.
- W4311463196 hasConceptScore W4311463196C187191949 @default.
- W4311463196 hasConceptScore W4311463196C199360897 @default.
- W4311463196 hasConceptScore W4311463196C2776214188 @default.
- W4311463196 hasConceptScore W4311463196C2777904410 @default.
- W4311463196 hasConceptScore W4311463196C2778476105 @default.
- W4311463196 hasConceptScore W4311463196C2780513914 @default.
- W4311463196 hasConceptScore W4311463196C41008148 @default.
- W4311463196 hasConceptScore W4311463196C50644808 @default.
- W4311463196 hasConceptScore W4311463196C78766204 @default.
- W4311463196 hasConceptScore W4311463196C80444323 @default.
- W4311463196 hasLocation W43114631961 @default.
- W4311463196 hasLocation W43114631962 @default.
- W4311463196 hasOpenAccess W4311463196 @default.
- W4311463196 hasPrimaryLocation W43114631961 @default.
- W4311463196 hasRelatedWork W1472213334 @default.
- W4311463196 hasRelatedWork W1998323989 @default.
- W4311463196 hasRelatedWork W2022059867 @default.
- W4311463196 hasRelatedWork W2086205578 @default.
- W4311463196 hasRelatedWork W2124870959 @default.
- W4311463196 hasRelatedWork W2151498641 @default.
- W4311463196 hasRelatedWork W2488897859 @default.
- W4311463196 hasRelatedWork W2900078434 @default.