IMC: Energy-eicient in-memory convolver for accelerating binarized deep neural network

Shaahin Angizi, Deliang Fan

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

4 Scopus citations

Abstract

Deep Convolutional Neural Networks (CNNs) are widely employed in modern AI systems due to their unprecedented accuracy in object recognition and detection. However, it has been proven that the main bottleneck to improve large scale deep CNN based hardware implementation performance is massive data communication between processing units and o-chip memory. In this paper, we pave a way towards novel concept of in-memory convolver (IMC) that could implement the dominant convolution computation within main memory based on our proposed Spin Orbit Torque Magnetic Random Access Memory (SOT-MRAM) array architecture to greatly reduce data communication and thus accelerate Binary CNN (BCNN). The proposed architecture could simultaneously work as non-volatile memory and a recongurable in-memory logic (AND, OR) without add-on logic circuits to memory chip as in conventional logic-in-memory designs. The computed logic output could be also simply read out like a normal MRAM bit-cell using the shared memory peripheral circuits. We employ such intrinsic in-memory processing architecture to eciently process data within memory to greatly reduce power-hungry and long distance data communication concerning state-of-the-art BCNN hardware. The hardware mapping results show that IMC can process the Binarized AlexNet on ImageNet data-set favorably with 134.27 J/img where ∼ 16× and 9× lower energy and area are achieved, respectively, compared to RRAM-based BCNN. Furthermore, 21.5% reduction in data movement in term of main memory accesses is observed compared to CPU/DRAM baseline.

Original languageEnglish (US)
Title of host publicationProceedings of Neuromorphic Computing Symposium, NCS 2017
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450364423
DOIs
StatePublished - Jul 17 2017
Externally publishedYes
Event2017 Neuromorphic Computing Symposium, NCS 2017 - Knoxville, United States
Duration: Jul 17 2017Jul 19 2017

Publication series

NameACM International Conference Proceeding Series
Volume2017-July

Conference

Conference2017 Neuromorphic Computing Symposium, NCS 2017
CountryUnited States
CityKnoxville
Period7/17/177/19/17

Keywords

  • Deep convolutional Neural Network
  • In-memory computing
  • Spin Hall eect

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'IMC: Energy-eicient in-memory convolver for accelerating binarized deep neural network'. Together they form a unique fingerprint.

Cite this