What To Know
- The automation tools for synthesizing a NASP chip from nearly any trained neural network dramatically reduce time and costs in chip design and allow the use of standard manufacturing methods, resulting in mass-market production cost savings.
- POLYN brings Tiny AI principles to neuromorphic processing of any sensor signal, handling analog or digital data and performing analog computations on the sensor level.
Itzhak Edrei has joined the company’s board of directors.
Itzhak Edrei will also assist POLYN in establishing strategic partnerships within the global semiconductor ecosystem.
“POLYN has attained technological achievements in the field of automation tools for superfast converting of nearly any neural network into chip layouts,” Edrei said. “That plus the company’s unique neuromorphic analog signal processing technology make a great foundation for the future on-device edge AI solutions, affordable for the mass market.”
“We’re proud to have Itzhak onboard as both his deep technology and business expertise will help the company move forward,” said Alexander Timofeev, CEO of POLYN. “With a track record of 30 years of experience in semiconductors, Itzhak’s expertise is comprehensive. We welcome the technological and operational excellence that Itzhak brings to our powerful team.”
Edrei is a veteran leader in the semiconductor industry, having held several high-level roles at Tower Semiconductor, a renowned global manufacturer of high value analog semiconductor solutions. He also now serves as Tower’s President Emeritus. Edrei holds a PhD in physics from Bar Ilan University and did post-doctoral research at Rutgers University.
POLYN brings Tiny AI principles to neuromorphic processing of any sensor signal, handling analog or digital data and performing analog computations on the sensor level. POLYN’s NASP technology provides sensor devices with true AI at unprecedentedly low power consumption, small size, and low latency.
POLYN offers best-in-class inference solutions for a wide range of sensors in Connected Health, IIoT, wearables, and other applications. Because NASP chips perform reliably even in the event of cloud connection failure, they ensure data safety and privacy. The automation tools for synthesizing a NASP chip from nearly any trained neural network dramatically reduce time and costs in chip design and allow the use of standard manufacturing methods, resulting in mass-market production cost savings.