Thursday, April 22, 2021

Expedera develops deep-learning accelerator (DLA) for AI silicon

Expedera, a start-up based in Santa Clara, California, emerged from stealth to unveil its Origin neural engine intellectual property (IP) for edge system silicon.

Expedera, which plans to license its deep-learning accelerator (DLA) technology to SoC designers, is targetting low-power edge devices like smartphones, tablets, computers, edge servers, and automotive. The company says its deep-learning accelerator (DLA) provides up to 18 TOPS/W at 7nm, which is up to ten times more than competitive offerings while minimizing memory requirements. Origin accelerates the performance of neural network models such as object detection, recognition, segmentation, super-resolution, and natural language processing. 

Expedera's co-founders include Da Chuang (CEO), who previously was cofounder and COO of Memoir Systems (an optimized memory IP startup acquired by Cisco); Siyad Ma (VP Engineering), who previously led Algorithmic TCAM ASIC and IP teams for Cisco Nexus7k, MDS, Cat4k/6k; and Sharad Chole (Chief Scientist), who previously was an architect at Cisco, Memoir Systems (Cisco), and Microsoft.