FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3141540147> ?p ?o ?g. }

• W3141540147 endingPage "738" @default.
• W3141540147 startingPage "730" @default.
• W3141540147 abstract "Deep neural networks (DNNs) have gained tremendous popularity in recent years due to their ability to achieve superhuman accuracy in a wide variety of machine learning tasks. However, the compute and memory requirements of DNNs have grown rapidly, creating a need for energy-efficient hardware. Resistive crossbars have attracted significant interest in the design of the next generation of DNN accelerators due to their ability to natively execute massively parallel vector-matrix multiplications within dense memory arrays. However, crossbar-based computations face a major challenge due to device and circuit-level nonidealities, which manifest as errors in the vector-matrix multiplications and eventually degrade DNN accuracy. To address this challenge, there is a need for tools that can model the functional impact of nonidealities on DNN training and inference. Existing efforts toward this goal are either limited to inference or are too slow to be used for large-scale DNN training. We propose TxSim, a fast and customizable modeling framework to functionally evaluate DNN training on crossbar-based hardware considering the impact of nonidealities. The key features of TxSim that differentiate it from prior efforts are: 1) it comprehensively models nonidealities during all training operations (forward propagation, backward propagation, and weight update) and 2) it achieves computational efficiency by mapping crossbar evaluations to well-optimized Basic Linear Algebra Subprograms (BLAS) routines and incorporates speedup techniques to further reduce simulation time with minimal impact on accuracy. TxSim achieves 6×- 108× improvement in simulation speed over prior works, and thereby makes it feasible to evaluate the training of large-scale DNNs on crossbars. Our experiments using TxSim reveal that the accuracy degradation in DNN training due to nonidealities can be substantial (3%-36.4%) for large-scale DNNs and data sets, underscoring the need for further research in mitigation techniques. We also analyze the impact of various device and circuit-level parameters and the associated nonidealities to provide key insights that can guide the design of crossbar-based DNN training accelerators." @default.
• W3141540147 created "2021-04-13" @default.
• W3141540147 creator A5065766721 @default.
• W3141540147 creator A5071950007 @default.
• W3141540147 creator A5083686495 @default.
• W3141540147 creator A5086674756 @default.
• W3141540147 date "2021-04-01" @default.
• W3141540147 modified "2023-10-06" @default.
• W3141540147 title "TxSim: Modeling Training of Deep Neural Networks on Resistive Crossbar Systems" @default.
• W3141540147 cites W2033811947 @default.
• W3141540147 cites W2060969833 @default.
• W3141540147 cites W2141879519 @default.
• W3141540147 cites W2307193480 @default.
• W3141540147 cites W2323986115 @default.
• W3141540147 cites W2545497739 @default.
• W3141540147 cites W2554279936 @default.
• W3141540147 cites W2606722458 @default.
• W3141540147 cites W2613989746 @default.
• W3141540147 cites W2782046614 @default.
• W3141540147 cites W2883929540 @default.
• W3141540147 cites W2946522000 @default.
• W3141540147 cites W2963387357 @default.
• W3141540147 cites W2980034233 @default.
• W3141540147 cites W2983353765 @default.
• W3141540147 cites W3031771821 @default.
• W3141540147 cites W3033519076 @default.
• W3141540147 cites W3106392217 @default.
• W3141540147 cites W3112740243 @default.
• W3141540147 cites W3142548120 @default.
• W3141540147 cites W4236709213 @default.
• W3141540147 cites W4242495425 @default.
• W3141540147 doi "https://doi.org/10.1109/tvlsi.2021.3063543" @default.
• W3141540147 hasPublicationYear "2021" @default.
• W3141540147 type Work @default.
• W3141540147 sameAs 3141540147 @default.
• W3141540147 citedByCount "22" @default.
• W3141540147 countsByYear W31415401472020 @default.
• W3141540147 countsByYear W31415401472021 @default.
• W3141540147 countsByYear W31415401472022 @default.
• W3141540147 countsByYear W31415401472023 @default.
• W3141540147 crossrefType "journal-article" @default.
• W3141540147 hasAuthorship W3141540147A5065766721 @default.
• W3141540147 hasAuthorship W3141540147A5071950007 @default.
• W3141540147 hasAuthorship W3141540147A5083686495 @default.
• W3141540147 hasAuthorship W3141540147A5086674756 @default.
• W3141540147 hasBestOaLocation W31415401472 @default.
• W3141540147 hasConcept C108583219 @default.
• W3141540147 hasConcept C113775141 @default.
• W3141540147 hasConcept C11413529 @default.
• W3141540147 hasConcept C118524514 @default.
• W3141540147 hasConcept C13164978 @default.
• W3141540147 hasConcept C154945302 @default.
• W3141540147 hasConcept C173608175 @default.
• W3141540147 hasConcept C2776214188 @default.
• W3141540147 hasConcept C29984679 @default.
• W3141540147 hasConcept C41008148 @default.
• W3141540147 hasConcept C42935608 @default.
• W3141540147 hasConcept C45374587 @default.
• W3141540147 hasConcept C50644808 @default.
• W3141540147 hasConcept C68339613 @default.
• W3141540147 hasConcept C76155785 @default.
• W3141540147 hasConcept C9390403 @default.
• W3141540147 hasConceptScore W3141540147C108583219 @default.
• W3141540147 hasConceptScore W3141540147C113775141 @default.
• W3141540147 hasConceptScore W3141540147C11413529 @default.
• W3141540147 hasConceptScore W3141540147C118524514 @default.
• W3141540147 hasConceptScore W3141540147C13164978 @default.
• W3141540147 hasConceptScore W3141540147C154945302 @default.
• W3141540147 hasConceptScore W3141540147C173608175 @default.
• W3141540147 hasConceptScore W3141540147C2776214188 @default.
• W3141540147 hasConceptScore W3141540147C29984679 @default.
• W3141540147 hasConceptScore W3141540147C41008148 @default.
• W3141540147 hasConceptScore W3141540147C42935608 @default.
• W3141540147 hasConceptScore W3141540147C45374587 @default.
• W3141540147 hasConceptScore W3141540147C50644808 @default.
• W3141540147 hasConceptScore W3141540147C68339613 @default.
• W3141540147 hasConceptScore W3141540147C76155785 @default.
• W3141540147 hasConceptScore W3141540147C9390403 @default.
• W3141540147 hasFunder F4320332180 @default.
• W3141540147 hasIssue "4" @default.
• W3141540147 hasLocation W31415401471 @default.
• W3141540147 hasLocation W31415401472 @default.
• W3141540147 hasOpenAccess W3141540147 @default.
• W3141540147 hasPrimaryLocation W31415401471 @default.
• W3141540147 hasRelatedWork W2799290373 @default.
• W3141540147 hasRelatedWork W2800886065 @default.
• W3141540147 hasRelatedWork W2910518169 @default.
• W3141540147 hasRelatedWork W2950973314 @default.
• W3141540147 hasRelatedWork W3141540147 @default.
• W3141540147 hasRelatedWork W3216454443 @default.
• W3141540147 hasRelatedWork W4280640895 @default.
• W3141540147 hasRelatedWork W4286233754 @default.
• W3141540147 hasRelatedWork W4288595854 @default.
• W3141540147 hasRelatedWork W4310614714 @default.
• W3141540147 hasVolume "29" @default.
• W3141540147 isParatext "false" @default.
• W3141540147 isRetracted "false" @default.
• W3141540147 magId "3141540147" @default.

• next