Matches in SemOpenAlex for { <https://semopenalex.org/work/W4224307323> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W4224307323 endingPage "3095" @default.
- W4224307323 startingPage "3095" @default.
- W4224307323 abstract "We propose a method for minimizing global buffer access within a deep learning accelerator for convolution operations by maximizing the data reuse through a local register file, thereby substituting the local register file access for the power-hungry global buffer access. To fully exploit the merits of data reuse, this study proposes a rearrangement of the computational sequence in a deep learning accelerator. Once input data are read from the global buffer, repeatedly reading the same data is performed only through the local register file, saving significant power consumption. Furthermore, different from prior works that equip local register files in each computation unit, the proposed method enables sharing a local register file along the column of the 2D computation array, saving resources and controlling overhead. The proposed accelerator is implemented on an off-the-shelf field-programmable gate array to verify the functionality and resource utilization. Then, the performance improvement of the proposed method is demonstrated relative to popular deep learning accelerators. Our evaluation indicates that the proposed deep learning accelerator reduces the number of global-buffer accesses to nearly 86.8%, consequently saving up to 72.3% of the power consumption for the input data memory access with a minor increase in resource usage compared to a conventional deep learning accelerator." @default.
- W4224307323 created "2022-04-26" @default.
- W4224307323 creator A5038525689 @default.
- W4224307323 creator A5078440061 @default.
- W4224307323 creator A5082407014 @default.
- W4224307323 creator A5091430365 @default.
- W4224307323 date "2022-04-18" @default.
- W4224307323 modified "2023-10-18" @default.
- W4224307323 title "Minimizing Global Buffer Access in a Deep Learning Accelerator Using a Local Register File with a Rearranged Computational Sequence" @default.
- W4224307323 cites W2289252105 @default.
- W4224307323 cites W2522820499 @default.
- W4224307323 cites W2604814848 @default.
- W4224307323 cites W2626712922 @default.
- W4224307323 cites W2795915628 @default.
- W4224307323 cites W2925935411 @default.
- W4224307323 cites W2945146780 @default.
- W4224307323 cites W2969414999 @default.
- W4224307323 cites W3040119159 @default.
- W4224307323 cites W3094627520 @default.
- W4224307323 doi "https://doi.org/10.3390/s22083095" @default.
- W4224307323 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35459079" @default.
- W4224307323 hasPublicationYear "2022" @default.
- W4224307323 type Work @default.
- W4224307323 citedByCount "1" @default.
- W4224307323 countsByYear W42243073232022 @default.
- W4224307323 crossrefType "journal-article" @default.
- W4224307323 hasAuthorship W4224307323A5038525689 @default.
- W4224307323 hasAuthorship W4224307323A5078440061 @default.
- W4224307323 hasAuthorship W4224307323A5082407014 @default.
- W4224307323 hasAuthorship W4224307323A5091430365 @default.
- W4224307323 hasBestOaLocation W42243073231 @default.
- W4224307323 hasConcept C108583219 @default.
- W4224307323 hasConcept C111919701 @default.
- W4224307323 hasConcept C117280010 @default.
- W4224307323 hasConcept C128916667 @default.
- W4224307323 hasConcept C153247305 @default.
- W4224307323 hasConcept C154945302 @default.
- W4224307323 hasConcept C165696696 @default.
- W4224307323 hasConcept C169590947 @default.
- W4224307323 hasConcept C173608175 @default.
- W4224307323 hasConcept C202491316 @default.
- W4224307323 hasConcept C2779960059 @default.
- W4224307323 hasConcept C2871975 @default.
- W4224307323 hasConcept C38652104 @default.
- W4224307323 hasConcept C41008148 @default.
- W4224307323 hasConcept C9390403 @default.
- W4224307323 hasConcept C98986596 @default.
- W4224307323 hasConceptScore W4224307323C108583219 @default.
- W4224307323 hasConceptScore W4224307323C111919701 @default.
- W4224307323 hasConceptScore W4224307323C117280010 @default.
- W4224307323 hasConceptScore W4224307323C128916667 @default.
- W4224307323 hasConceptScore W4224307323C153247305 @default.
- W4224307323 hasConceptScore W4224307323C154945302 @default.
- W4224307323 hasConceptScore W4224307323C165696696 @default.
- W4224307323 hasConceptScore W4224307323C169590947 @default.
- W4224307323 hasConceptScore W4224307323C173608175 @default.
- W4224307323 hasConceptScore W4224307323C202491316 @default.
- W4224307323 hasConceptScore W4224307323C2779960059 @default.
- W4224307323 hasConceptScore W4224307323C2871975 @default.
- W4224307323 hasConceptScore W4224307323C38652104 @default.
- W4224307323 hasConceptScore W4224307323C41008148 @default.
- W4224307323 hasConceptScore W4224307323C9390403 @default.
- W4224307323 hasConceptScore W4224307323C98986596 @default.
- W4224307323 hasIssue "8" @default.
- W4224307323 hasLocation W42243073231 @default.
- W4224307323 hasLocation W42243073232 @default.
- W4224307323 hasLocation W42243073233 @default.
- W4224307323 hasLocation W42243073234 @default.
- W4224307323 hasOpenAccess W4224307323 @default.
- W4224307323 hasPrimaryLocation W42243073231 @default.
- W4224307323 hasRelatedWork W1551927567 @default.
- W4224307323 hasRelatedWork W1861891114 @default.
- W4224307323 hasRelatedWork W2034841960 @default.
- W4224307323 hasRelatedWork W2053999332 @default.
- W4224307323 hasRelatedWork W2122846563 @default.
- W4224307323 hasRelatedWork W2128881050 @default.
- W4224307323 hasRelatedWork W3041671716 @default.
- W4224307323 hasRelatedWork W4248626492 @default.
- W4224307323 hasRelatedWork W2059399861 @default.
- W4224307323 hasRelatedWork W2463380484 @default.
- W4224307323 hasVolume "22" @default.
- W4224307323 isParatext "false" @default.
- W4224307323 isRetracted "false" @default.
- W4224307323 workType "article" @default.