Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384111957> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4384111957 abstract "Deep neural networks (DNNs) are of critical use in different domains. To accelerate DNN computation, tensor compilers are proposed to generate efficient code on different domain-specific accelerators. Existing tensor compilers mainly focus on optimizing computation efficiency. However, memory access is becoming a key performance bottleneck because the computational performance of accelerators is increasing much faster than memory performance. The lack of direct description of memory access and data dependence in current tensor compilers' intermediate representation (IR) brings significant challenges to generate memory-efficient code. In this paper, we propose IntelliGen, a tensor compiler that can generate high-performance code for memory-intensive operators by considering both computation and data movement optimizations. IntelliGen represent a DNN program using GIR, which includes primitives indicating its computation, data movement, and parallel strategies. This information will be further composed as an instruction-level dataflow graph to perform holistic optimizations by searching different memory access patterns and computation operations, and generating memory-efficient code on different hardware. We evaluate IntelliGen on NVIDIA GPU, AMD GPU, and Cambricon MLU, showing speedup up to 1.97x, 2.93x, and 16.91x(1.28x, 1.23x, and 2.31x on average), respectively, compared to current most performant frameworks." @default.
- W4384111957 created "2023-07-13" @default.
- W4384111957 creator A5019792492 @default.
- W4384111957 creator A5023674199 @default.
- W4384111957 creator A5023681514 @default.
- W4384111957 creator A5024162917 @default.
- W4384111957 creator A5031342822 @default.
- W4384111957 creator A5039766485 @default.
- W4384111957 creator A5050585748 @default.
- W4384111957 creator A5051544755 @default.
- W4384111957 creator A5066504632 @default.
- W4384111957 creator A5071200777 @default.
- W4384111957 date "2023-07-10" @default.
- W4384111957 modified "2023-10-17" @default.
- W4384111957 title "PowerFusion: A Tensor Compiler with Explicit Data Movement Description and Instruction-level Graph IR" @default.
- W4384111957 doi "https://doi.org/10.48550/arxiv.2307.04995" @default.
- W4384111957 hasPublicationYear "2023" @default.
- W4384111957 type Work @default.
- W4384111957 citedByCount "0" @default.
- W4384111957 crossrefType "posted-content" @default.
- W4384111957 hasAuthorship W4384111957A5019792492 @default.
- W4384111957 hasAuthorship W4384111957A5023674199 @default.
- W4384111957 hasAuthorship W4384111957A5023681514 @default.
- W4384111957 hasAuthorship W4384111957A5024162917 @default.
- W4384111957 hasAuthorship W4384111957A5031342822 @default.
- W4384111957 hasAuthorship W4384111957A5039766485 @default.
- W4384111957 hasAuthorship W4384111957A5050585748 @default.
- W4384111957 hasAuthorship W4384111957A5051544755 @default.
- W4384111957 hasAuthorship W4384111957A5066504632 @default.
- W4384111957 hasAuthorship W4384111957A5071200777 @default.
- W4384111957 hasBestOaLocation W43841119571 @default.
- W4384111957 hasConcept C132525143 @default.
- W4384111957 hasConcept C133162039 @default.
- W4384111957 hasConcept C149635348 @default.
- W4384111957 hasConcept C169590947 @default.
- W4384111957 hasConcept C173608175 @default.
- W4384111957 hasConcept C190902152 @default.
- W4384111957 hasConcept C199360897 @default.
- W4384111957 hasConcept C26517878 @default.
- W4384111957 hasConcept C2780513914 @default.
- W4384111957 hasConcept C38652104 @default.
- W4384111957 hasConcept C41008148 @default.
- W4384111957 hasConcept C45374587 @default.
- W4384111957 hasConcept C68339613 @default.
- W4384111957 hasConcept C77660490 @default.
- W4384111957 hasConcept C80444323 @default.
- W4384111957 hasConcept C96324660 @default.
- W4384111957 hasConceptScore W4384111957C132525143 @default.
- W4384111957 hasConceptScore W4384111957C133162039 @default.
- W4384111957 hasConceptScore W4384111957C149635348 @default.
- W4384111957 hasConceptScore W4384111957C169590947 @default.
- W4384111957 hasConceptScore W4384111957C173608175 @default.
- W4384111957 hasConceptScore W4384111957C190902152 @default.
- W4384111957 hasConceptScore W4384111957C199360897 @default.
- W4384111957 hasConceptScore W4384111957C26517878 @default.
- W4384111957 hasConceptScore W4384111957C2780513914 @default.
- W4384111957 hasConceptScore W4384111957C38652104 @default.
- W4384111957 hasConceptScore W4384111957C41008148 @default.
- W4384111957 hasConceptScore W4384111957C45374587 @default.
- W4384111957 hasConceptScore W4384111957C68339613 @default.
- W4384111957 hasConceptScore W4384111957C77660490 @default.
- W4384111957 hasConceptScore W4384111957C80444323 @default.
- W4384111957 hasConceptScore W4384111957C96324660 @default.
- W4384111957 hasLocation W43841119571 @default.
- W4384111957 hasOpenAccess W4384111957 @default.
- W4384111957 hasPrimaryLocation W43841119571 @default.
- W4384111957 hasRelatedWork W1547259518 @default.
- W4384111957 hasRelatedWork W1583465708 @default.
- W4384111957 hasRelatedWork W1814870153 @default.
- W4384111957 hasRelatedWork W184060744 @default.
- W4384111957 hasRelatedWork W2069811640 @default.
- W4384111957 hasRelatedWork W2134386444 @default.
- W4384111957 hasRelatedWork W2569816949 @default.
- W4384111957 hasRelatedWork W4200123712 @default.
- W4384111957 hasRelatedWork W4285049632 @default.
- W4384111957 hasRelatedWork W2479014312 @default.
- W4384111957 isParatext "false" @default.
- W4384111957 isRetracted "false" @default.
- W4384111957 workType "article" @default.