FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3102587717> ?p ?o ?g. }

• W3102587717 abstract "TensorDash is a hardware level technique for enabling data-parallel MAC units to take advantage of sparsity in their input operand streams. When used to compose a hardware accelerator for deep learning, TensorDash can speedup the training process while also increasing energy efficiency. TensorDash combines a low-cost, sparse input operand interconnect comprising an 8-input multiplexer per multiplier input, with an area-efficient hardware scheduler. While the interconnect allows a very limited set of movements per operand, the scheduler can effectively extract sparsity when it is present in the activations, weights or gradients of neural networks. Over a wide set of models covering various applications, TensorDash accelerates the training process by $1.95{times}$ while being $1.89times$ more energy-efficient, $1.6times$ more energy efficient when taking on-chip and off-chip memory accesses into account. While TensorDash works with any datatype, we demonstrate it with both single-precision floating-point units and bfloat16." @default.
• W3102587717 created "2020-11-23" @default.
• W3102587717 creator A5000665425 @default.
• W3102587717 creator A5005950098 @default.
• W3102587717 creator A5007585346 @default.
• W3102587717 creator A5023853738 @default.
• W3102587717 creator A5045620647 @default.
• W3102587717 creator A5072544251 @default.
• W3102587717 creator A5090867030 @default.
• W3102587717 date "2020-10-01" @default.
• W3102587717 modified "2023-10-16" @default.
• W3102587717 title "TensorDash: Exploiting Sparsity to Accelerate Deep Neural Network Training" @default.
• W3102587717 cites W1840435438 @default.
• W3102587717 cites W1861492603 @default.
• W3102587717 cites W2048266589 @default.
• W3102587717 cites W2194775991 @default.
• W3102587717 cites W2289252105 @default.
• W3102587717 cites W2442974303 @default.
• W3102587717 cites W2463955103 @default.
• W3102587717 cites W2515287984 @default.
• W3102587717 cites W2516141709 @default.
• W3102587717 cites W2565639579 @default.
• W3102587717 cites W2605350416 @default.
• W3102587717 cites W2606722458 @default.
• W3102587717 cites W2625457103 @default.
• W3102587717 cites W2626211758 @default.
• W3102587717 cites W2626991402 @default.
• W3102587717 cites W2794952988 @default.
• W3102587717 cites W2884150179 @default.
• W3102587717 cites W2904902077 @default.
• W3102587717 cites W2931118404 @default.
• W3102587717 cites W2962821792 @default.
• W3102587717 cites W2963150697 @default.
• W3102587717 cites W2963566954 @default.
• W3102587717 cites W2963594949 @default.
• W3102587717 cites W2963809228 @default.
• W3102587717 cites W2964194679 @default.
• W3102587717 cites W2979310060 @default.
• W3102587717 cites W2979439447 @default.
• W3102587717 cites W2981474208 @default.
• W3102587717 cites W2982219368 @default.
• W3102587717 cites W2982236015 @default.
• W3102587717 cites W2984140583 @default.
• W3102587717 cites W3004495293 @default.
• W3102587717 cites W3102587717 @default.
• W3102587717 cites W4240168186 @default.
• W3102587717 cites W624975885 @default.
• W3102587717 doi "https://doi.org/10.1109/micro50266.2020.00069" @default.
• W3102587717 hasPublicationYear "2020" @default.
• W3102587717 type Work @default.
• W3102587717 sameAs 3102587717 @default.
• W3102587717 citedByCount "38" @default.
• W3102587717 countsByYear W31025877172020 @default.
• W3102587717 countsByYear W31025877172021 @default.
• W3102587717 countsByYear W31025877172022 @default.
• W3102587717 countsByYear W31025877172023 @default.
• W3102587717 crossrefType "proceedings-article" @default.
• W3102587717 hasAuthorship W3102587717A5000665425 @default.
• W3102587717 hasAuthorship W3102587717A5005950098 @default.
• W3102587717 hasAuthorship W3102587717A5007585346 @default.
• W3102587717 hasAuthorship W3102587717A5023853738 @default.
• W3102587717 hasAuthorship W3102587717A5045620647 @default.
• W3102587717 hasAuthorship W3102587717A5072544251 @default.
• W3102587717 hasAuthorship W3102587717A5090867030 @default.
• W3102587717 hasBestOaLocation W31025877172 @default.
• W3102587717 hasConcept C105795698 @default.
• W3102587717 hasConcept C111919701 @default.
• W3102587717 hasConcept C11413529 @default.
• W3102587717 hasConcept C118524514 @default.
• W3102587717 hasConcept C119599485 @default.
• W3102587717 hasConcept C123745756 @default.
• W3102587717 hasConcept C124584101 @default.
• W3102587717 hasConcept C127413603 @default.
• W3102587717 hasConcept C139719470 @default.
• W3102587717 hasConcept C154945302 @default.
• W3102587717 hasConcept C162324750 @default.
• W3102587717 hasConcept C165005293 @default.
• W3102587717 hasConcept C173608175 @default.
• W3102587717 hasConcept C177264268 @default.
• W3102587717 hasConcept C186370098 @default.
• W3102587717 hasConcept C19275194 @default.
• W3102587717 hasConcept C199360897 @default.
• W3102587717 hasConcept C2742236 @default.
• W3102587717 hasConcept C31258907 @default.
• W3102587717 hasConcept C33923547 @default.
• W3102587717 hasConcept C41008148 @default.
• W3102587717 hasConcept C50644808 @default.
• W3102587717 hasConcept C55526617 @default.
• W3102587717 hasConcept C68339613 @default.
• W3102587717 hasConcept C70970002 @default.
• W3102587717 hasConcept C76155785 @default.
• W3102587717 hasConcept C84211073 @default.
• W3102587717 hasConcept C9390403 @default.
• W3102587717 hasConcept C98045186 @default.
• W3102587717 hasConceptScore W3102587717C105795698 @default.
• W3102587717 hasConceptScore W3102587717C111919701 @default.
• W3102587717 hasConceptScore W3102587717C11413529 @default.
• W3102587717 hasConceptScore W3102587717C118524514 @default.
• W3102587717 hasConceptScore W3102587717C119599485 @default.
• W3102587717 hasConceptScore W3102587717C123745756 @default.

• next