Toronto-based artificial intelligence (AI) chip startup company Untether AI has emerged from stealth mode and announced it raised $13 million in Series A funding led by Intel Capital along with a number of other unnamed investors. Founded by CEO Martin Snelgrove, CTO Darrick Wiebe, and Raymond Chik last year, Untether AI is developing a new AI chip that eliminates the data movement bottleneck that costs energy and performance in traditional architectures. And the team behind the company has experience in delivering more than 1 billion chips to market.
“We’re at a historic moment in computing right now. AI is changing the nature of what computers do. They can now interact with the real world based on recognizing patterns instead of following procedures. These advances in AI have opened up tremendous possibilities for business and technological innovation, but traditional processors are not best suited to this new type of computing. Now is the time to build a new kind of ultra-efficient chip for new frontiers in AI applications — from autonomous vehicles, data centers, mobile, vision processing and beyond — that will ultimately change how everything works,” said Snelgrove in a statement.
Untether AI’s groundbreaking new chip architecture for neural net inference combines the power efficiency of near-memory design with the robustness of digital processing. And it reduces the distance data must travel to the absolute minimum. Untether AI’s Kensington architecture gets data to the processors at 2.5 petabits per second at a rate 1,000 times better than what is possible in traditional architectures. This eliminates the cost of using a bus to move data to processors — which traditionally dominates energy consumption. By improving inference efficiency, Untether AI enables larger scale use and broad artificial intelligence while consuming fewer resources and energy.
“In under five months after a small initial seed investment, Untether AI was able to build a successful prototype based on a processing-near-memory architecture that promises to enable dramatically higher efficiencies and performance for AI accelerators. This is a team with decades of experience in chip design, and they’re now developing a commercial product for neural net inference – in an AI inference market that’s projected to grow to $22 billion by 2022,” added Intel Capital VP and senior managing director Dave Flanagan.
Wiebe explained that the AI inference market is poised for disruption and rapid growth. And by introducing a completely scalable and ultra-efficient inference architecture, Untether AI is “uniquely positioned to address each of those market segments with what will be by far the most energy-efficient AI accelerator.