Convolutional Networks for Fast, Energy-Efficient Neuromorphic Computing

TL;DR

This paper demonstrates that neuromorphic hardware can efficiently implement deep convolutional networks that achieve near state-of-the-art accuracy on multiple datasets while maintaining high energy efficiency and throughput.

Contribution

It shows how to implement and train deep convolutional networks on neuromorphic chips, combining deep learning accuracy with neuromorphic energy efficiency.

Findings

Achieves near state-of-the-art accuracy on 8 datasets

Runs at 1200-2600 frames/sec with 25-275 mW power

Enables training of neuromorphic networks using backpropagation

Abstract

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that i) approach state-of-the-art classification accuracy across 8 standard datasets, encompassing vision and speech, ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1200 and 2600 frames per second and using between 25 and 275 mW (effectively > 6000 frames / sec / W) and iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. For the first time, the algorithmic power of deep learning can be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.