Matches in SemOpenAlex for { <https://semopenalex.org/work/W4322576240> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W4322576240 endingPage "14" @default.
- W4322576240 startingPage "9" @default.
- W4322576240 abstract "Across different Deep Learning (DL) applications or within the same application but in different phases, bitwidth precision of activations and weights may vary. Moreover, energy and latency of MAC units have to be minimized, especially at the edge. Hence, various precision-scalable MAC units optimized for DL have recently emerged. Our contribution is a new precision-configurable multiplier/dot-product unit based on a modified Radix-4 Booth signed multiplier with Sum-Together (ST) mode. Besides 16-bit full precision multiplications, it can be reconfigured to perform dot products among two 8-bit or four 4-bit sub words of the input operands without requiring an external adder, thus reducing the number of cycles of MAC operations. The results of the synthesis in performance, power and area on a 28-nm technology show that our unit (1) is superior to other state of the art ST multipliers in area ( $$approx $$ 35% less) in the clock frequency range between 100 and 1000 MHz and (2) reduces latency up to 4x when used to compute a convolutional layer, at the cost of limited overheads in area (+10%) and power (+13%) compared to a conventional 16-bit Booth multiplier. This unit can play an important role in designing variable-precision MAC units or DL accelerators for edge devices." @default.
- W4322576240 created "2023-02-28" @default.
- W4322576240 creator A5036528695 @default.
- W4322576240 creator A5067223738 @default.
- W4322576240 date "2023-01-01" @default.
- W4322576240 modified "2023-09-29" @default.
- W4322576240 title "A Reconfigurable Multiplier/Dot-Product Unit for Precision-Scalable Deep Learning Applications" @default.
- W4322576240 cites W2141697220 @default.
- W4322576240 cites W2162098969 @default.
- W4322576240 cites W2963255460 @default.
- W4322576240 cites W2963367920 @default.
- W4322576240 cites W2987956272 @default.
- W4322576240 cites W4312997782 @default.
- W4322576240 doi "https://doi.org/10.1007/978-3-031-26066-7_2" @default.
- W4322576240 hasPublicationYear "2023" @default.
- W4322576240 type Work @default.
- W4322576240 citedByCount "0" @default.
- W4322576240 crossrefType "book-chapter" @default.
- W4322576240 hasAuthorship W4322576240A5036528695 @default.
- W4322576240 hasAuthorship W4322576240A5067223738 @default.
- W4322576240 hasConcept C124584101 @default.
- W4322576240 hasConcept C139719470 @default.
- W4322576240 hasConcept C162324750 @default.
- W4322576240 hasConcept C164620267 @default.
- W4322576240 hasConcept C187919765 @default.
- W4322576240 hasConcept C2524010 @default.
- W4322576240 hasConcept C32900221 @default.
- W4322576240 hasConcept C33652231 @default.
- W4322576240 hasConcept C33923547 @default.
- W4322576240 hasConcept C41008148 @default.
- W4322576240 hasConcept C48044578 @default.
- W4322576240 hasConcept C55526617 @default.
- W4322576240 hasConcept C72475854 @default.
- W4322576240 hasConcept C76155785 @default.
- W4322576240 hasConcept C77088390 @default.
- W4322576240 hasConcept C82876162 @default.
- W4322576240 hasConcept C9390403 @default.
- W4322576240 hasConcept C94375191 @default.
- W4322576240 hasConceptScore W4322576240C124584101 @default.
- W4322576240 hasConceptScore W4322576240C139719470 @default.
- W4322576240 hasConceptScore W4322576240C162324750 @default.
- W4322576240 hasConceptScore W4322576240C164620267 @default.
- W4322576240 hasConceptScore W4322576240C187919765 @default.
- W4322576240 hasConceptScore W4322576240C2524010 @default.
- W4322576240 hasConceptScore W4322576240C32900221 @default.
- W4322576240 hasConceptScore W4322576240C33652231 @default.
- W4322576240 hasConceptScore W4322576240C33923547 @default.
- W4322576240 hasConceptScore W4322576240C41008148 @default.
- W4322576240 hasConceptScore W4322576240C48044578 @default.
- W4322576240 hasConceptScore W4322576240C55526617 @default.
- W4322576240 hasConceptScore W4322576240C72475854 @default.
- W4322576240 hasConceptScore W4322576240C76155785 @default.
- W4322576240 hasConceptScore W4322576240C77088390 @default.
- W4322576240 hasConceptScore W4322576240C82876162 @default.
- W4322576240 hasConceptScore W4322576240C9390403 @default.
- W4322576240 hasConceptScore W4322576240C94375191 @default.
- W4322576240 hasLocation W43225762401 @default.
- W4322576240 hasOpenAccess W4322576240 @default.
- W4322576240 hasPrimaryLocation W43225762401 @default.
- W4322576240 hasRelatedWork W17194118 @default.
- W4322576240 hasRelatedWork W1976891713 @default.
- W4322576240 hasRelatedWork W2185364297 @default.
- W4322576240 hasRelatedWork W2350334566 @default.
- W4322576240 hasRelatedWork W2803703254 @default.
- W4322576240 hasRelatedWork W3113088580 @default.
- W4322576240 hasRelatedWork W4322576240 @default.
- W4322576240 hasRelatedWork W80620896 @default.
- W4322576240 hasRelatedWork W2184742891 @default.
- W4322576240 hasRelatedWork W2359786673 @default.
- W4322576240 isParatext "false" @default.
- W4322576240 isRetracted "false" @default.
- W4322576240 workType "book-chapter" @default.