A New Learning Method for Inference Accuracy, Core Occupation, and Performance Co-optimization on TrueNorth Chip

04/03/2016
by   Wei Wen, et al.
0

IBM TrueNorth chip uses digital spikes to perform neuromorphic computing and achieves ultrahigh execution parallelism and power efficiency. However, in TrueNorth chip, low quantization resolution of the synaptic weights and spikes significantly limits the inference (e.g., classification) accuracy of the deployed neural network model. Existing workaround, i.e., averaging the results over multiple copies instantiated in spatial and temporal domains, rapidly exhausts the hardware resources and slows down the computation. In this work, we propose a novel learning method on TrueNorth platform that constrains the random variance of each computation copy and reduces the number of needed copies. Compared to the existing learning method, our method can achieve up to 68.8 slightly improved inference accuracy.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 3

page 4

01/07/2017

Classification Accuracy Improvement for Neuromorphic Computing Systems with One-level Precision Synapses

Brain inspired neuromorphic computing has demonstrated remarkable advant...
12/16/2016

Neuromorphic Deep Learning Machines

An ongoing challenge in neuromorphic computing is to devise general and ...
12/21/2020

Efficient On-Chip Learning for Optical Neural Networks Through Power-Aware Sparse Zeroth-Order Optimization

Optical neural networks (ONNs) have demonstrated record-breaking potenti...
11/18/2019

Efficient Hardware Implementation of Incremental Learning and Inference on Chip

In this paper, we tackle the problem of incrementally learning a classif...
05/08/2020

GOBO: Quantizing Attention-Based NLP Models for Low Latency and Energy Efficient Inference

Attention-based models have demonstrated remarkable success in various n...
04/13/2021

Mitigating Adversarial Attack for Compute-in-Memory Accelerator Utilizing On-chip Finetune

Compute-in-memory (CIM) has been proposed to accelerate the convolution ...
02/24/2018

Privacy Leakages in Approximate Adders

Approximate computing has recently emerged as a promising method to meet...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.