FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3091813304> ?p ?o ?g. }

• W3091813304 abstract "Conventional neural accelerators rely on isolated self-sufficient functional units that perform an atomic operation while communicating the results through an operand delivery-aggregation logic. Each single unit processes all the bits of their operands atomically and produce all the bits of the results in isolation. This paper explores a different design style, where each unit is only responsible for a slice of the bit-level operations to interleave and combine the benefits of bit-level parallelism with the abundant data-level parallelism in deep neural networks. A dynamic collection of these units cooperate at runtime to generate bits of the results, collectively. Such cooperation requires extracting new grouping between the bits, which is only possible if the operands and operations are vectorizable. The abundance of Data-Level Parallelism and mostly repeated execution patterns, provides a unique opportunity to define and leverage this new dimension of Bit-Parallel Vector Composability. This design intersperses bit parallelism within data-level parallelism and dynamically interweaves the two together. As such, the building block of our neural accelerator is a Composable Vector Unit that is a collection of Narrower-Bitwidth Vector Engines, which are dynamically composed or decomposed at the bit granularity. Using six diverse CNN and LSTM deep networks, we evaluate this design style across four design points: with and without algorithmic bitwidth heterogeneity and with and without availability of a high-bandwidth off-chip memory. Across these four design points, Bit-Parallel Vector Composability brings (1.4× to 3.5×) speedup and (1.1× to 2.7×) energy reduction. We also comprehensively compare our design style to the Nvidia's RTX 2080 TI GPU, which also supports INT-4 execution. The benefits range between 28.0× and 33.7× improvement in Performance-per-Watt." @default.
• W3091813304 created "2020-10-15" @default.
• W3091813304 creator A5035427716 @default.
• W3091813304 creator A5037648751 @default.
• W3091813304 creator A5082318034 @default.
• W3091813304 creator A5082499242 @default.
• W3091813304 creator A5084514143 @default.
• W3091813304 date "2020-07-01" @default.
• W3091813304 modified "2023-10-12" @default.
• W3091813304 title "Bit-Parallel Vector Composability for Neural Acceleration" @default.
• W3091813304 cites W2285660444 @default.
• W3091813304 cites W2442974303 @default.
• W3091813304 cites W2516141709 @default.
• W3091813304 cites W2518281301 @default.
• W3091813304 cites W2524428287 @default.
• W3091813304 cites W2563587242 @default.
• W3091813304 cites W2606722458 @default.
• W3091813304 cites W2625457103 @default.
• W3091813304 cites W2719597717 @default.
• W3091813304 cites W2751477244 @default.
• W3091813304 cites W2761132374 @default.
• W3091813304 cites W2786771851 @default.
• W3091813304 cites W2794141774 @default.
• W3091813304 cites W2883929540 @default.
• W3091813304 cites W2909365574 @default.
• W3091813304 cites W2946572707 @default.
• W3091813304 cites W2949870694 @default.
• W3091813304 cites W2963367920 @default.
• W3091813304 doi "https://doi.org/10.1109/dac18072.2020.9218656" @default.
• W3091813304 hasPublicationYear "2020" @default.
• W3091813304 type Work @default.
• W3091813304 sameAs 3091813304 @default.
• W3091813304 citedByCount "10" @default.
• W3091813304 countsByYear W30918133042020 @default.
• W3091813304 countsByYear W30918133042021 @default.
• W3091813304 countsByYear W30918133042022 @default.
• W3091813304 countsByYear W30918133042023 @default.
• W3091813304 crossrefType "proceedings-article" @default.
• W3091813304 hasAuthorship W3091813304A5035427716 @default.
• W3091813304 hasAuthorship W3091813304A5037648751 @default.
• W3091813304 hasAuthorship W3091813304A5082318034 @default.
• W3091813304 hasAuthorship W3091813304A5082499242 @default.
• W3091813304 hasAuthorship W3091813304A5084514143 @default.
• W3091813304 hasBestOaLocation W30918133042 @default.
• W3091813304 hasConcept C120314980 @default.
• W3091813304 hasConcept C154945302 @default.
• W3091813304 hasConcept C173608175 @default.
• W3091813304 hasConcept C2778814252 @default.
• W3091813304 hasConcept C2781172179 @default.
• W3091813304 hasConcept C41008148 @default.
• W3091813304 hasConcept C50644808 @default.
• W3091813304 hasConcept C55526617 @default.
• W3091813304 hasConcept C61483411 @default.
• W3091813304 hasConcept C68339613 @default.
• W3091813304 hasConcept C80444323 @default.
• W3091813304 hasConcept C9390403 @default.
• W3091813304 hasConceptScore W3091813304C120314980 @default.
• W3091813304 hasConceptScore W3091813304C154945302 @default.
• W3091813304 hasConceptScore W3091813304C173608175 @default.
• W3091813304 hasConceptScore W3091813304C2778814252 @default.
• W3091813304 hasConceptScore W3091813304C2781172179 @default.
• W3091813304 hasConceptScore W3091813304C41008148 @default.
• W3091813304 hasConceptScore W3091813304C50644808 @default.
• W3091813304 hasConceptScore W3091813304C55526617 @default.
• W3091813304 hasConceptScore W3091813304C61483411 @default.
• W3091813304 hasConceptScore W3091813304C68339613 @default.
• W3091813304 hasConceptScore W3091813304C80444323 @default.
• W3091813304 hasConceptScore W3091813304C9390403 @default.
• W3091813304 hasLocation W30918133041 @default.
• W3091813304 hasLocation W30918133042 @default.
• W3091813304 hasOpenAccess W3091813304 @default.
• W3091813304 hasPrimaryLocation W30918133041 @default.
• W3091813304 hasRelatedWork W1531488649 @default.
• W3091813304 hasRelatedWork W1585350690 @default.
• W3091813304 hasRelatedWork W1608806855 @default.
• W3091813304 hasRelatedWork W1685698149 @default.
• W3091813304 hasRelatedWork W2023505575 @default.
• W3091813304 hasRelatedWork W2074226157 @default.
• W3091813304 hasRelatedWork W2209942166 @default.
• W3091813304 hasRelatedWork W2313503008 @default.
• W3091813304 hasRelatedWork W2366027386 @default.
• W3091813304 hasRelatedWork W4249523967 @default.
• W3091813304 isParatext "false" @default.
• W3091813304 isRetracted "false" @default.
• W3091813304 magId "3091813304" @default.
• W3091813304 workType "article" @default.