PUFsecurity, Himax team up to take user and data security to new heights
PUFsecurity and Himax Technologies have collaborated to successfully integrate PUFsecurity’s PUF-based Root of Trust IP, PUFrt, into Himax’s WiseEye2 AI processor (WE2), a low-power AI chip. This integration aims to transform endpoint AI application performance while prioritizing user security and data protection.
PUFsecurity’s Hardware Root of Trust solution (PUFrt) offers a security solution for AI models and user data protection. It also offers features such as 1024-bit identification code (PUF), zero-touch key provisioning, and secure One-Time Programmable (OTP) storage, boot and updates.
According to Mark Chen, the smart sensing business vice president at Himax, the WiseEyeTM AI processors are known for their industrial-grade security, cryptography engines and low power consumption compared to other AI processors. He continues that the collaboration with PUFsecurity enhances WE2 capabilities in security, privacy, battery life, responsiveness, and applicability across domains like consumer electronics, laptops, surveillance, medical and agriculture.
Chairperson and president of PUFsecurity, Dr. Charles Hsu, also comments: “On top of safeguarding the AI model from being altered or stolen, it also empowers the entire system’s continuous secure operations with very efficient power consumption.”
Additionally, an Arm-commissioned report emphasized user preferences for longer battery life, faster system response and privacy in public spaces. Himax says WE2 addresses these concerns by incorporating a multi-layer power management architecture, dual CPU cores and a microNPU core for efficient neural network operations at the edge. Moreover, the company notes that WE2 ensures privacy with PUFsecurity, safeguarding AI models and data at the hardware level.
Last year, Lenovo selected Sequitur Labs, an edge security firm, to protect edge AI computing applications built on the ThinkEdge SE70.
Sequitur Labs’ EmSPARK Security Suite addresses embedded security concerns, focusing on protecting AI models at the edge.
The companies note that it provides encryption, storage, data transmission and key/certificate management within a microprocessor’s secure memory partition. Combined with the Lenovo ThinkEdge SE70 edge AI platform, the companies say it enables secure isolation of dedicated hardware for running AI models and delivering inferences and data.
Read more:
Axelera AI secures a total of $50 million in funding in 20 months to develop edge AI chips
BrainChip, SiFive put heads together, work on deploying next-gen edge AI chips
Navigating the edge: A recap of key edge computing developments in 2023
Article Topics
AI chip | AI processing | edge computing | Himax Technologies | PUFsecurity
Comments