IP partners to develop security-conscious AI platforms

September 05, 2019 //By Julien Happich
AI platforms
A provider of security IP cores, Silex Insight has partnered with AI accelerator solutions provider Wave Computing to deliver a combined solution for security-conscious semiconductor designers incorporating IP from both Wave (MIPS) and Silex Insight.

AI systems require a safe and secure platform for updating AI algorithms, AI models and securing data that has privacy concerns. These proliferating security-conscious applications are especially important to the enterprise and automotive markets, which include all devices that are network connected.

The partnership provides a combined platform comprised of Wave’s MIPS Open IP (or MIPS commercial CPU IP) platforms, along with Silex Insight’s automotive-grade security eSecure IP platform. 

  • eSecure acts as a programmable Root-of-Trust. At its center is a secure controller that takes full control over all security functions, especially those that shield the system’s secrets such as encryption keys from the main applications running on the product. The module also contains a scalable cryptographic engine supporting symmetric and asymmetric operations, secure debugging capability, authentication of components, firmware and application software, advanced anti-tampering features, protection against side-channel attacks, a secure storage interface, and much more.

  • MIPS Open is the first free, silicon-proven architecture available for open use without license fees. It offers a unified approach and is based on mature MIPS CPU technology. SoC designers can also easily upgrade from MIPS Open IP to the full-featured, high-performance, commercial MIPS processor IP platform, which is automotive- qualified. Billions of devices have been shipped based on proven MIPS IP. 

Silex Insight - www.silexinsight.com
Wave Computing - https://wavecomp.ai


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.