BrainChip licenses unsupervised learning for SNAP

Technology News |
By Peter Clarke

JAST is an unsupervised learning algorithm developed by Simon Thorpe and his team at Cerco and is a complementary approach to traditional deep learning techniques.

The addition of the JAST technology to BrainChip’s Spiking Neuron Adaptive Processor (SNAP) will provide a compact and easy to use alternative to the use of GPU processors for visual analytics.

“The JAST technology adds to our offering of SNvision, a software solution that runs on any Windows platform, and our hardware solution, SNAPvision, which will be available in the third quarter of 2017,” said Louis DiNardo, CEO of BrainChip, in a statement. “The ability to learn on the device, and at near instant speeds, sets our technology apart from other solutions that require long training cycles. The JAST learning rules will provide an exciting enhancement to our existing product offerings,” he added.

The terms of the exclusive license have not been disclosed.

Related links and articles:

News articles:

BrainChip buys French computer vision firm

BrainChip appoints former Exar CEO to lead company

ARM’s soft launch for machine learning library

EE Times Silicon 60: 2016’s Emerging Companies to Watch


Linked Articles
eeNews Europe