Abstract
Spiking Neural Networks (SNNs) represent a biologically inspired computation model capable of emulating neural computation in human brain and brain-like structures. The main promise is very low energy consumption. Classic Von Neumann architecture based SNN accelerators in hardware, however, often fall short of addressing demanding computation and data transfer requirements efficiently at scale. In this article, we propose a promising alternative to overcome scalability limitations, based on a network of in-memory SNN accelerators, which can reduce the energy consumption by up to 150.25= when compared to a representative ASIC solution. The significant reduction in energy comes from two key aspects of the hardware design to minimize data communication overheads: (1) each node represents an in-memory SNN accelerator based on a spintronic Computational RAM array, and (2) a novel, De Bruijn graph based architecture establishes the SNN array connectivity.
| Original language | English (US) |
|---|---|
| Article number | 3475963 |
| Journal | ACM Transactions on Architecture and Code Optimization |
| Volume | 18 |
| Issue number | 4 |
| DOIs | |
| State | Published - Dec 2021 |
Bibliographical note
Funding Information:This work was supported in part by NSF grant no. SPX-1725420. Authors. address: H. Cılasun, S. Resch, Z. I. Chowdhury, E. Olson, M. Zabihi, Z. Zhao, T. Peterson, K. K. Parhi, J.-P. Wang, S. S. Sapatnekar, and U. R. Karpuzcu, University of Minnesota, Twin Cities, 200 Union St SE, Minneapolis, Minnesota, 55455; emails: {cilas001, resc0059, chowh005, olso6834, zabih003, zhaox526, pete9290, parhi, jpwang, sachin, ukarpuzc}@umn.edu. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from permissions@acm.org. © 2021 Association for Computing Machinery. 1544-3566/2021/09-ART59 $15.00 https://doi.org/10.1145/3475963
Publisher Copyright:
© 2021 ACM.
Keywords
- Processing in memory
- computational random access memory
- non-volatile memory
- spiking neural networks