Industry change

OpenAI models advance chip design

Published
Source
OpenAI blog

Summary

OpenAI shared results from applying its models to semiconductor design beyond its in-house inference chip Jalapeno. By the company's account, its model improved power, performance, and area by up to 14% on AMD's flagship GPU IP relative to heavily optimized baselines, an estimated year's worth of progress in under five months, and some changes are already in designs sent for manufacturing. AMD also verified single-digit gains on key blocks of its next flagship CPU, and scores on open-source RISC-V cores rose with each model generation. OpenAI is now training a GPT-Synopsys model with Synopsys to use EDA tools the way expert designers do.

Why it matters for our work

The work shows how far models can take over expert-level design tasks, with human-reviewed changes reaching production chip designs. Improvement figures and timelines are claims by OpenAI and AMD.

Translated from the Korean original. Summaries may be translated and edited. Commentary reflects our perspective; forecasts remain the source’s views.

Read original (opens in a new tab)