Presentation 2004/11/20
Implementation of 1000 neuron hardware system based on stochastic logic and it's application
Akiyoshi MOMOI, Shigeo SATO, Koji NAKAJIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We have built a 1000 neuron hardware system based on stochastic logic. Stochastic logic realizes pseudo analog operations using stochastically coded digital pulses. Therefore, with using stochastic logic we can realize a smaller circuit than binary logic and higher reliability than an analog circuit. Furthermore, we can control the magnitude of the coding noise by changing the accumulation time, in order to enhance the performance. We report detail and some measurement results of our system.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Neural network / Hardware / Stochastic operation / Noise
Paper # NLP2004-82,NC2004-98
Date of Issue

Conference Information
Committee NC
Conference Date 2004/11/20(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Implementation of 1000 neuron hardware system based on stochastic logic and it's application
Sub Title (in English)
Keyword(1) Neural network
Keyword(2) Hardware
Keyword(3) Stochastic operation
Keyword(4) Noise
1st Author's Name Akiyoshi MOMOI
1st Author's Affiliation Graduate School of Information Science, Tohoku University()
2nd Author's Name Shigeo SATO
2nd Author's Affiliation Laboratory for Brainware Systems, Laboratory for Nanoelectronics and Spintronics, Research Institute of Electrical Communication, Tohoku University
3rd Author's Name Koji NAKAJIMA
3rd Author's Affiliation Laboratory for Brainware Systems, Laboratory for Nanoelectronics and Spintronics, Research Institute of Electrical Communication, Tohoku University
Date 2004/11/20
Paper # NLP2004-82,NC2004-98
Volume (vol) vol.104
Number (no) 474
Page pp.pp.-
#Pages 6
Date of Issue