Minerva: Enabling Low-Power, Highly-Accurate Deep Neural Network Accelerators

Brandon Reagen, Paul Whatmough, Robert Adolf, Saketh Rama, Hyunkwang Lee, Sae Kyu Lee, Jose Miguel Hernandez-Lobato, Gu Yeon Wei, David Brooks

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The continued success of Deep Neural Networks (DNNs) in classification tasks has sparked a trend of accelerating their execution with specialized hardware. While published designs easily give an order of magnitude improvement over general-purpose hardware, few look beyond an initial implementation. This paper presents Minerva, a highly automated co-design approach across the algorithm, architecture, and circuit levels to optimize DNN hardware accelerators. Compared to an established fixed-point accelerator baseline, we show that fine-grained, heterogeneous data type optimization reduces power by 1.5, aggressive, in-line predication and pruning of small activity values further reduces power by 2.0, and active hardware fault detection coupled with domain-aware error mitigation eliminates an additional 2.7 through lowering SRAM voltages. Across five datasets, these optimizations provide a collective average of 8.1 power reduction over an accelerator baseline without compromising DNN model accuracy. Minerva enables highly accurate, ultra-low power DNN accelerators (in the range of tens of milliwatts), making it feasible to deploy DNNs in power-constrained IoT and mobile devices.

Original languageEnglish (US)
Title of host publicationProceedings - 2016 43rd International Symposium on Computer Architecture, ISCA 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages267-278
Number of pages12
ISBN (Electronic)9781467389471
DOIs
StatePublished - Aug 24 2016
Event43rd International Symposium on Computer Architecture, ISCA 2016 - Seoul, Korea, Republic of
Duration: Jun 18 2016Jun 22 2016

Publication series

NameProceedings - 2016 43rd International Symposium on Computer Architecture, ISCA 2016

Conference

Conference43rd International Symposium on Computer Architecture, ISCA 2016
CountryKorea, Republic of
CitySeoul
Period6/18/166/22/16

ASJC Scopus subject areas

  • Hardware and Architecture

Fingerprint Dive into the research topics of 'Minerva: Enabling Low-Power, Highly-Accurate Deep Neural Network Accelerators'. Together they form a unique fingerprint.

Cite this