A Neural Network Decomposition Algorithm for Mapping on Crossbar-Based Computing Systems
Open Access
- 18 September 2020
- journal article
- research article
- Published by MDPI AG in Electronics
- Vol. 9 (9) , 1526
Abstract
Crossbar-based neuromorphic computing to accelerate neural networks is a popular alternative to conventional von Neumann computing systems. It is also referred as processing-in-memory and in-situ analog computing. The crossbars have a fixed number of synapses per neuron and it is necessary to decompose neurons to map networks onto the crossbars. This paper proposes the k-spare decomposition algorithm that can trade off the predictive performance against the neuron usage during the mapping. The proposed algorithm performs a two-level hierarchical decomposition. In the first global decomposition, it decomposes the neural network such that each crossbar has k spare neurons. These neurons are used to improve the accuracy of the partially mapped network in the subsequent local decomposition. Our experimental results using modern convolutional neural networks show that the proposed method can improve the accuracy substantially within about 10% extra neurons.Keywords
Funding Information
- National Research Foundation of Korea (NRF-2017R1D1A1B03029103)
- Institute for Information and Communications Technology Promotion (1711073912)
This publication has 15 references indexed in Scilit:
- Efficient Hardware Architectures for Deep Convolutional Neural NetworkIEEE Transactions on Circuits and Systems I: Regular Papers, 2017
- A Tutorial on Fisher informationJournal of Mathematical Psychology, 2017
- ISAACACM SIGARCH Computer Architecture News, 2016
- PRIMEACM SIGARCH Computer Architecture News, 2016
- Mastering the game of Go with deep neural networks and tree searchNature, 2016
- TrueNorth: Design and Tool Flow of a 65 mW 1 Million Neuron Programmable Neurosynaptic ChipIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2015
- Deep learningNature, 2015
- The processor-memory bottleneckXRDS: Crossroads, The ACM Magazine for Students, 1999
- The PIM architecture for wide-area multicast routingIEEE/ACM Transactions on Networking, 1996
- Neuromorphic electronic systemsProceedings of the IEEE, 1990