OpenAI announced that its AI models collaborated with AMD to achieve up to 14% improvement in power, performance, and area (PPA) metrics on the chipmaker's flagship GPU IP. The optimization was measured against an already well-optimized baseline, with further gains identified by rewriting circuit description code—a process OpenAI claims compressed roughly a year's worth of engineering progress into under five months. AMD has integrated some of these AI-generated changes into designs recently sent to manufacturing, while remaining improvements have been incorporated into production design files for future versions. Additionally, AMD verified that model-generated designs delivered single-digit percentage PPA improvements on key modules of its next-generation flagship CPU. The same methodology was also applied to Jalapeño, OpenAI's first in-house inference chip, with related improvements confirmed on actual silicon.