1.7pJ/SOP Neuromorphic Processor with Integrated Partial Sum Routers for In-Network Computing

1.7pJ/SOP Neuromorphic Processor with Integrated Partial Sum Routers for In-Network Computing

2023 IEEE International Symposium on Circuits and Systems (ISCAS)

10.1109/ISCAS46773.2023.10181759

http://dx.doi.org/10.1109/iscas46773.2023.10181759

B. Wang, M. M. Wong, D. Li, Y. S. Chong, J. Zhou, W. F. Wong, L. Peh, A. Mani, M. Upadhyay, A. Balaji, A. T. Do

21 July 2023

Wang, B., Wong, M. M., Li, D., Chong, Y. S., Zhou, J., Wong, W. F., Peh, L., Mani, A., Upadhyay, M., Balaji, A., & Do, A. T. (2023, May 21). 1.7pJ/SOP Neuromorphic Processor with Integrated Partial Sum Routers for In-Network Computing. 2023 IEEE International Symposium on Circuits and Systems (ISCAS). https://doi.org/10.1109/iscas46773.2023.10181759

Conventional neuromorphic accelerators primarily leverage split-merge method to accommodate a neural network that is beyond a single core's size, leading to possible accuracy loss, extra core usage and significant power and energy overhead. This work presents an energy-efficient, reconfigurable neuro-morphic processor to address the problem by (i) a partial sum router circuitry that enables in-network computing to remove the need of extra merge cores; (ii) software-defined Networks-on-Chip that eliminates the power-hungry routing compute and (iii) fine-grained power gating and clock gating technique for power reduction. Our test chip achieves lossless mapping as the algorithm and an energy efficiency of 1.7pJ/SOP at 0.5V, 19% lower than state-of-the-art result.

Publisher Copyright

This research / project is supported by the National Research Foundation - Competitive Research Programme
Grant Reference no. : NRF-CRP23- 2019-0003

This research / project is supported by the Singapore University of Technology and Design (SUTD) - MOE AcRF Tier-1 Grant via SUTD Start-Up Research Grant
Grant Reference no. : SRT3IS20162

This research / project is supported by the Singapore University of Technology and Design (SUTD) - MOE AcRF Tier-1 Grant via SUTD Growth Plan Grant – AI Sector
Grant Reference no. : SGP-AIRS1841

This research / project is supported by the Singapore University of Technology and Design (SUTD) - MOE AcRF Tier-1 Grant via SUTD-ZJU IDEA Visiting Professors Grant
Grant Reference no. : SUTD-ZJU(VP)201808

© 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

https://oar.a-star.edu.sg/communities-collections/articles/20267

2158-1525

Institute of Microelectronics