Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Park, Gunho | - |
dc.contributor.author | Kung, Jaeha | - |
dc.contributor.author | Lee, Youngjoo | - |
dc.date.accessioned | 2021-10-07T12:00:05Z | - |
dc.date.available | 2021-10-07T12:00:05Z | - |
dc.date.created | 2021-05-14 | - |
dc.date.issued | 2021-07 | - |
dc.identifier.issn | 1549-8328 | - |
dc.identifier.uri | http://hdl.handle.net/20.500.11750/15432 | - |
dc.description.abstract | In this paper, we present a novel approximate computing scheme suitable for realizing the energy-efficient multiply-accumulate (MAC) processing. In contrast to the prior works that suffer from the error accumulation limiting the approximate range, we utilize different approximate multipliers in an interleaved way to compensate errors in the opposite direction during accumulate operations. For the balanced error accumulation, we first design the approximate 4-2 compressors generating errors in the opposite direction while minimizing the computational costs. Based on the probabilistic analysis, positive and negative multipliers are then carefully developed to provide a similar error distance. Simulation results on various practical applications reveal that the proposed MAC processing offers the energy-efficient computing scenario by extending the range of approximate parts. Even compared to the state-of-the-art solutions, for example, the proposed interleaving scheme relaxes the core-level energy consumption of the recent CNN accelerator by more than 35% without degrading the recognition accuracy. IEEE | - |
dc.language | English | - |
dc.publisher | Institute of Electrical and Electronics Engineers | - |
dc.title | Design and Analysis of Approximate Compressors for Balanced Error Accumulation in MAC Operator | - |
dc.type | Article | - |
dc.identifier.doi | 10.1109/TCSI.2021.3073177 | - |
dc.identifier.scopusid | 2-s2.0-85105030224 | - |
dc.identifier.bibliographicCitation | IEEE Transactions on Circuits and Systems I: Regular Papers, v.68, no.7, pp.2950 - 2961 | - |
dc.description.isOpenAccess | FALSE | - |
dc.subject.keywordAuthor | Approximate computing | - |
dc.subject.keywordAuthor | Approximation algorithms | - |
dc.subject.keywordAuthor | Compressors | - |
dc.subject.keywordAuthor | convolutional neural network | - |
dc.subject.keywordAuthor | Energy consumption | - |
dc.subject.keywordAuthor | Hardware | - |
dc.subject.keywordAuthor | image processing | - |
dc.subject.keywordAuthor | low-power circuit design | - |
dc.subject.keywordAuthor | Mathematical model | - |
dc.subject.keywordAuthor | multiplier. | - |
dc.subject.keywordAuthor | Signal processing algorithms | - |
dc.subject.keywordPlus | Energy efficiency | - |
dc.subject.keywordPlus | Energy utilization | - |
dc.subject.keywordPlus | Computational costs | - |
dc.subject.keywordPlus | Design and analysis | - |
dc.subject.keywordPlus | Energy efficient | - |
dc.subject.keywordPlus | Energy efficient computing | - |
dc.subject.keywordPlus | Error accumulation | - |
dc.subject.keywordPlus | Multiply accumulate | - |
dc.subject.keywordPlus | Probabilistic analysis | - |
dc.subject.keywordPlus | Recognition accuracy | - |
dc.subject.keywordPlus | Errors | - |
dc.citation.endPage | 2961 | - |
dc.citation.number | 7 | - |
dc.citation.startPage | 2950 | - |
dc.citation.title | IEEE Transactions on Circuits and Systems I: Regular Papers | - |
dc.citation.volume | 68 | - |
There are no files associated with this item.