Minimizing area and energy of deep learning hardware design using collective low precision and structured compression

Shihui Yin, Gaurav Srivastava, Shreyas K. Venkataramanaiah, Chaitali Chakrabarti, Visar Berisha, Jae-sun Seo

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

5 Scopus citations

Abstract

Deep learning algorithms have shown tremendous success in many recognition tasks; however, these algorithms typically include a deep neural network (DNN) structure and a large number of parameters, which makes it challenging to implement them on power/area-constrained embedded platforms. To reduce the network size, several studies investigated compression by introducing element-wise or row-/column-/block-wise sparsity via pruning and regularization. In addition, many recent works have focused on reducing precision of activations and weights with some reducing down to a single bit. However, combining various sparsity structures with binarized or very-low-precision (2-3 bit) neural networks have not been comprehensively explored. In this work, we present design techniques for minimum-area/-energy DNN hardware with minimal degradation in accuracy. During training, both binarization/low-precision and structured sparsity are applied as constraints to find the smallest memory footprint for a given deep learning algorithm. The DNN model for CIFAR-10 dataset with weight memory reduction of 50X exhibits accuracy comparable to that of the floating-point counterpart. Area, performance and energy results of DNN hardware in 40nm CMOS are reported for the MNIST dataset. The optimized DNN that combines 8X structured compression and 3-bit weight precision showed 98.4% accuracy at 20nJ per classification.

Original languageEnglish (US)
Title of host publicationConference Record of 51st Asilomar Conference on Signals, Systems and Computers, ACSSC 2017
EditorsMichael B. Matthews
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1907-1911
Number of pages5
ISBN (Electronic)9781538618233
DOIs
StatePublished - Apr 10 2018
Event51st Asilomar Conference on Signals, Systems and Computers, ACSSC 2017 - Pacific Grove, United States
Duration: Oct 29 2017Nov 1 2017

Publication series

NameConference Record of 51st Asilomar Conference on Signals, Systems and Computers, ACSSC 2017
Volume2017-October

Other

Other51st Asilomar Conference on Signals, Systems and Computers, ACSSC 2017
Country/TerritoryUnited States
CityPacific Grove
Period10/29/1711/1/17

ASJC Scopus subject areas

  • Control and Optimization
  • Computer Networks and Communications
  • Hardware and Architecture
  • Signal Processing
  • Biomedical Engineering
  • Instrumentation

Fingerprint

Dive into the research topics of 'Minimizing area and energy of deep learning hardware design using collective low precision and structured compression'. Together they form a unique fingerprint.

Cite this