comment on this article

US company unveils first commercial spiking neural network architecture

BrainChip, has become the first company to bring a production spiking neural network architecture – the Akida Neuromorphic System-on-Chip (NSoC) – to market.

The Akida NSoC is intended for use in edge applications such as advanced driver assistance systems (ADAS), autonomous vehicles, drones, vision-guided robotics, surveillance and machine vision systems.

The architecture is scalable which, according to the company, will allow users to network Akida devices together to perform complex neural network training and inferencing.

“The artificial intelligence acceleration chipset marketplace is expected to surpass $60billion by 2025,” said Aditya Kaul, Research Director at Tractica, a leading market intelligence firm with a specialisation in AI. “Neuromorphic computing holds significant promise to accelerate AI, especially for low-power applications.”

The Akida NSoC uses a pure CMOS logic process and spiking neural networks (SNNs) are inherently lower power than traditional convolutional neural networks (CNNs), replacing math-intensive convolutions and back-propagation training methods with biologically inspired neuron functions and feed-forward training methodologies.

The Akida NSoC is designed for use as a stand-alone embedded accelerator or as a co-processor. It includes sensor interfaces for traditional pixel-based imaging, dynamic vision sensors (DVS), Lidar, audio, and analogue signals. It also has high-speed data interfaces. Embedded in the NSoC are data-to-spike converters designed to optimally convert popular data formats into spikes to train and be processed by the Akida Neuron Fabric.

Author
Bethan Grylls

Comment on this article


This material is protected by MA Business copyright See Terms and Conditions. One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.

What you think about this article:

Spiking neural nets are the future of artificial intelligence. They are based on pulse timing (just like the brain) and they use a lot less power than conventional neural nets. They will make deep neural nets obsolete sooner or later.

Posted by: Louis Savain, 10/09/2018

Add your comments

Name
 
Email
 
Comments
 

Your comments/feedback may be edited prior to publishing. Not all entries will be published.
Please view our Terms and Conditions before leaving a comment.

Related Articles