The layout for this high-performance, ultra-low power chip was achieved in partnership with Socionext. Socionext has released the engineering layout of the production version of the AKD1000 chip to TSMC, which has begun preparing to manufacture at volume.
“I am grateful to our engineering team, who worked hard over the past eight months to release the Akida technology for volume production, and to our EAP customers that have helped lead us to market readiness,” said Peter van der Made, BrainChip CEO. “This move to manufacturing is a major milestone for BrainChip and for the industry at large as the first realistic opportunity to bring AI processing capability to edge devices for learning, enabling personalization of products without the need for retraining.”
BrainChip Early Access Program (EAP) customers purchased Akida evaluation systems for a range of edge applications and following testing of the AKD1000 engineering samples, BrainChip was able to improve the design in terms of performance, efficiency, and scalability, adding additional operating modes for even lower power consumption than its original design.
The Akida (NSoC) and intellectual property, can be used in applications including smart home, smart health, smart city and smart transportation.
Akida has been designed for use as a stand-alone embedded accelerator or as a co-processor, and includes interfaces for ADAS sensors, audio sensors, and other IoT sensors.
Production units are expected to be available around August 2021.
