Who Founded Cerebras?

Andrew Feldman co-founded Cerebras Systems in 2015. After selling his previous server company to AMD, he recognized that traditional CPUs and GPUs were fundamentally bottlenecked by data transfer speeds when training massive AI models. As CEO, he championed the audacious, highly contrarian idea of building a 'Wafer-Scale Engine' (WSE)—a single, massive computer chip the size of an entire silicon wafer—to completely eliminate communication bottlenecks between processors. (Exact birth year remains publicly undisclosed).

Gary Lauterbach co-founded Cerebras Systems. Widely considered one of the most brilliant silicon architects in the world, he solved the massive, decades-old physical challenges of wafer-scale integration—specifically figuring out how to bypass inevitable manufacturing defects and deliver power across a massive, unbroken 8.5-by-8.5 inch square of silicon. Gary Lauterbach was born in 1961.

Sean Lie is a co-founder and Chief Hardware Architect at Cerebras Systems. A prodigy in digital logic design, Lie architected the internal 2D mesh fabric of the Wafer-Scale Engine, connecting 900,000 independent AI processor cores with line-rate, low-latency inter-core communication. Lie invented the hardware-level defect tolerance algorithms that allow the WSE-3 to automatically route around flawed silicon dies during power-up, solving the yield problem that had baffled the global semiconductor industry for over four decades. Lie's architectural skill was central to overcoming the 50-year-old 'wafer yield impossibility' that destroyed previous semiconductor pioneers. By conceiving a hardware-level defect tolerance system featuring redundant cores and a reconfigurable 2D mesh, Lie ensured that microscopic silicon defects do not destroy the wafer. At Cerebras, Lie directs overall silicon microarchitecture, instruction set design, and next-generation packaging technologies with TSMC. (Exact birth year remains publicly undisclosed).

Michael James co-founded Cerebras Systems alongside his former SeaMicro colleagues. As Chief Architect of Advanced Technologies, he was responsible for designing the highly specialized compiler software required to take standard AI models (written in PyTorch or TensorFlow) and successfully distribute them across the millions of cores on the massive Cerebras Wafer-Scale Engine. (Exact birth year remains publicly undisclosed).

Jean-Philippe Fricker is a highly accomplished systems engineer and a co-founder of Cerebras Systems. Like the rest of the Cerebras founding team, Fricker was a key, foundational member of SeaMicro, where he served as a senior hardware engineer designing high-density server architectures. When the team departed AMD to launch Cerebras in 2016, Fricker took on the immensely complex task of engineering the physical systems, power delivery, and cooling infrastructure required to house the Wafer-Scale Engine. Because the WSE generates thousands of watts of heat across a single piece of silicon, Fricker's highly innovative system-level engineering was absolutely critical in transforming the theoretical wafer into a functional, deployable AI supercomputer. (Exact birth year remains publicly undisclosed).