AI · November 18, 2026 · Moderate impact
OpenAI says its models delivered a year of chip optimization on AMD's flagship GPU IP in under five months
Working with AMD, OpenAI achieved up to a 14 percent PPA (power, performance, area) improvement on AMD's flagship GPU IP in less than five months, which OpenAI says is roughly a year's worth of progress. AMD has already folded the changes into a design sent to production. Separately, OpenAI says it used GPT-6 Astra to optimize circuits used in Jalapeño silicon and to improve RISC-V core designs. Anthropic is also applying AI to chip work to improve cost and performance per dollar.
It shows AI models producing measurable gains in hardware design that have reached a production chip, not just benchmarks.