San Francisco-based startup Wafer AI claims its AI-driven optimization can get AMD’s MI355X GPU to 80% of Nvidia B200 throughput at less than half the cost. In July 2026, Wafer fine-tuned Z.AI’s GLM-5.2 model on AMD’s MI355X GPU. The company closed a $40 million Series A funding round in September 2026, valuing it at over $200 million, following a $4 million seed round in April 2026. Investors include Google Jeff Dean and OpenAI’s Wojciech Zaremba. Co-founders Emilio Andere and Steven Arellano, both University of Chicago alumni, built AI agents that autonomously optimize GPU workloads, delivering 2 to 2.8x speedups over stock baselines.
Source: Read the original article
