iCreatAI News
Skip to content
The Decoder· Matthias Bastian·· 2 days agoFeaturedAI score78

OpenAI says its models delivered a year's worth of chip optimization on top of AMD's own engineering

OpenAI says its models delivered a year's worth of chip optimization on top of AMD's own engineering

AI summary

Working with AMD, OpenAI achieved up to 14 percent PPA improvement on AMD's flagship GPU IP in less than five months, which OpenAI says equals about a year's progress on top of designs AMD engineers had already optimized for years, with some changes already folded into a design sent to manufacturing. The same method helped with Jalapeño, OpenAI's first custom inference chip, and in tests with open RISC-V processor cores the GPT-6.2 Astra model delivered the best results across model generations. OpenAI is also working with Synopsys on GPT-Synopsys to make AI-assisted chip design more broadly accessible, while Anthropic's Mythos model handled prompt processing for a model on AMD MI355X GPUs in three days instead of the six to twelve months Anthropic says the task normally takes.

Why it matters

The piece offers a concrete look at how AI models are being applied to chip design at AMD, OpenAI and Anthropic, with specific figures and timelines.

Source: The Decoder · the-decoder.com