Matches in SemOpenAlex for { <https://semopenalex.org/work/W3200747779> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W3200747779 abstract "Spiking neural networks (SNNs) have emerged as a new generation of neural networks, presenting a brain-inspired event-driven model with advantages in spatiotemporal information processing. Due to the need for high power consumption of compute-intensive neural accelerators, adequate power delivery network (PDN) design is a key requirement to ensure power efficiency and integrity. However, PDN design for SNN accelerators has not been extensively studied despite its great potential benefit in energy efficiency. In this paper, we present the first study on dynamic heterogeneous voltage regulation (HVR) for spiking neural accelerators to maximize system energy efficiency while ensuring power integrity. We propose a novel sparse-workload-aware dynamic PDN control policy, which enables high energy efficiency of sparse spiking computation on a systolic array. By exploring sparse inputs and all-or-none nature of spiking computations for PDN control, we explore different types of PDNs to accelerate spiking convolutional neural networks (S-CNNs) trained with the dynamic vision sensor (DVS) gesture dataset. Furthermore, we demonstrate various power gating schemes to further optimize the proposed PDN architecture, which leads to a more than a three-fold reduction in total energy overhead for spiking neural computations on systolic array-based accelerators." @default.
- W3200747779 created "2021-09-27" @default.
- W3200747779 creator A5037151690 @default.
- W3200747779 creator A5048357268 @default.
- W3200747779 creator A5070891492 @default.
- W3200747779 creator A5071458554 @default.
- W3200747779 date "2021-07-18" @default.
- W3200747779 modified "2023-10-16" @default.
- W3200747779 title "Systolic-Array Spiking Neural Accelerators with Dynamic Heterogeneous Voltage Regulation" @default.
- W3200747779 cites W1604973310 @default.
- W3200747779 cites W2044535169 @default.
- W3200747779 cites W2067523571 @default.
- W3200747779 cites W2069851236 @default.
- W3200747779 cites W2085332361 @default.
- W3200747779 cites W2109596721 @default.
- W3200747779 cites W2114601852 @default.
- W3200747779 cites W2117696986 @default.
- W3200747779 cites W2152839228 @default.
- W3200747779 cites W2289252105 @default.
- W3200747779 cites W2513853720 @default.
- W3200747779 cites W2612055717 @default.
- W3200747779 cites W2745933219 @default.
- W3200747779 cites W2783525259 @default.
- W3200747779 cites W2912311811 @default.
- W3200747779 cites W2961245123 @default.
- W3200747779 cites W3028593627 @default.
- W3200747779 cites W3041495229 @default.
- W3200747779 cites W3146763006 @default.
- W3200747779 cites W4238614602 @default.
- W3200747779 doi "https://doi.org/10.1109/ijcnn52387.2021.9534037" @default.
- W3200747779 hasPublicationYear "2021" @default.
- W3200747779 type Work @default.
- W3200747779 sameAs 3200747779 @default.
- W3200747779 citedByCount "2" @default.
- W3200747779 countsByYear W32007477792023 @default.
- W3200747779 crossrefType "proceedings-article" @default.
- W3200747779 hasAuthorship W3200747779A5037151690 @default.
- W3200747779 hasAuthorship W3200747779A5048357268 @default.
- W3200747779 hasAuthorship W3200747779A5070891492 @default.
- W3200747779 hasAuthorship W3200747779A5071458554 @default.
- W3200747779 hasConcept C111919701 @default.
- W3200747779 hasConcept C11413529 @default.
- W3200747779 hasConcept C11731999 @default.
- W3200747779 hasConcept C119599485 @default.
- W3200747779 hasConcept C127413603 @default.
- W3200747779 hasConcept C14580979 @default.
- W3200747779 hasConcept C149635348 @default.
- W3200747779 hasConcept C150741067 @default.
- W3200747779 hasConcept C154945302 @default.
- W3200747779 hasConcept C2742236 @default.
- W3200747779 hasConcept C2778476105 @default.
- W3200747779 hasConcept C2779960059 @default.
- W3200747779 hasConcept C2780165032 @default.
- W3200747779 hasConcept C41008148 @default.
- W3200747779 hasConcept C45374587 @default.
- W3200747779 hasConcept C50644808 @default.
- W3200747779 hasConcept C81363708 @default.
- W3200747779 hasConceptScore W3200747779C111919701 @default.
- W3200747779 hasConceptScore W3200747779C11413529 @default.
- W3200747779 hasConceptScore W3200747779C11731999 @default.
- W3200747779 hasConceptScore W3200747779C119599485 @default.
- W3200747779 hasConceptScore W3200747779C127413603 @default.
- W3200747779 hasConceptScore W3200747779C14580979 @default.
- W3200747779 hasConceptScore W3200747779C149635348 @default.
- W3200747779 hasConceptScore W3200747779C150741067 @default.
- W3200747779 hasConceptScore W3200747779C154945302 @default.
- W3200747779 hasConceptScore W3200747779C2742236 @default.
- W3200747779 hasConceptScore W3200747779C2778476105 @default.
- W3200747779 hasConceptScore W3200747779C2779960059 @default.
- W3200747779 hasConceptScore W3200747779C2780165032 @default.
- W3200747779 hasConceptScore W3200747779C41008148 @default.
- W3200747779 hasConceptScore W3200747779C45374587 @default.
- W3200747779 hasConceptScore W3200747779C50644808 @default.
- W3200747779 hasConceptScore W3200747779C81363708 @default.
- W3200747779 hasFunder F4320306076 @default.
- W3200747779 hasLocation W32007477791 @default.
- W3200747779 hasOpenAccess W3200747779 @default.
- W3200747779 hasPrimaryLocation W32007477791 @default.
- W3200747779 hasRelatedWork W1481681735 @default.
- W3200747779 hasRelatedWork W1697832939 @default.
- W3200747779 hasRelatedWork W1983101679 @default.
- W3200747779 hasRelatedWork W2100799217 @default.
- W3200747779 hasRelatedWork W2233116163 @default.
- W3200747779 hasRelatedWork W2367396372 @default.
- W3200747779 hasRelatedWork W2567624757 @default.
- W3200747779 hasRelatedWork W2949961122 @default.
- W3200747779 hasRelatedWork W3134760666 @default.
- W3200747779 hasRelatedWork W4293023302 @default.
- W3200747779 isParatext "false" @default.
- W3200747779 isRetracted "false" @default.
- W3200747779 magId "3200747779" @default.
- W3200747779 workType "article" @default.