On November 18 OpenAI published results from using its frontier models for chip design. Working with AMD, its model rewrote SystemVerilog RTL and found up to 14% PPA (power, performance, area) improvement on AMD's heavily optimized flagship GPU IP in under five months, and AMD has already put some changes into a design sent for manufacturing. On OpenAI's own Jalapeño inference chip, model-proposed changes cut standard-cell area by about 13% in the warp engine and 10% in the matrix unit, proven in silicon. OpenAI also shared RISC-V core evals where GPT-6.2 Astra scored highest, and said it is training a new model, GPT-Synopsys, with Synopsys.

Key Takeaways

  • ✓AMD: up to 14% PPA improvement on the flagship GPU IP versus heavily optimized baselines, which OpenAI calls roughly a year of progress in under five months; some changes are already in a design AMD sent for manufacturing
  • ✓AMD also verified single-digit percentage PPA gains across key blocks of its upcoming next-generation flagship CPU
  • ✓Jalapeño: model-proposed changes cut standard-cell area by about 13% in the warp engine and 10% in the matrix unit; one multiplier used about 15% less area while dropping latency from two cycles to one, all silicon proven
  • ✓RISC-V evals: on the open-source PicoRV32 and Ibex cores, GPT-6.2 Astra had the highest average score at about 112,000 tokens per run, across 9,881 scored runs
  • ✓Synopsys: the companies are training GPT-Synopsys to use Synopsys EDA tools like expert chip designers; rewrites were checked for functional equivalence with Synopsys tools
OpenAI Shows Its Models Designing Chips: Up to 14% PPA Gain on AMD Flagship GPU IP, 10–13% Area Savings on Jalapeño, and GPT-Synopsys Built with Synopsys
🖼️Official Media
Click to view high-res
🧭

Turn your technical choice into a development budget

Compare 40 dev plans & simulate token costs vs $20/mo subscriptions

🔬

In-Depth Technical Analysis

Background

RTL optimization depends on senior microarchitects, mature IP blocks are often tuned for years, and every few points of PPA (power, performance, area) are hard won. In August OpenAI said its models had sped up development of Jalapeño, its first custom inference chip. This November 18 blog post details those gains and shows results on third-party hardware from AMD for the first time.

How it works

OpenAI says the model reasons about RTL like an expert microarchitect and rewrites circuits that compute, control operations or move data. Each rewrite was checked with Synopsys tools to confirm the block's behavior was preserved and to weigh PPA trade-offs. On Jalapeño, models have helped since GPT-4o in early 2025, with gains in floating-point adders, multipliers, dot-product logic and pipeline staging. Next, OpenAI and Synopsys are training GPT-Synopsys so the model can drive EDA tools directly.

Numbers that matter

Up to 14% PPA improvement on AMD's flagship GPU IP and single-digit percentage gains on key blocks of its next-generation flagship CPU. On Jalapeño, standard-cell area fell about 13% in the warp engine and 10% in the matrix unit. In RISC-V evals on PicoRV32 and Ibex, GPT-6.2 Astra had the highest average score at about 112,000 tokens per run, roughly twice GPT-6 Astra on PicoRV32. All figures are OpenAI's own, with no independent reproduction yet.

What it means for developers

This is a research and partnership result, not a product you can call today, and OpenAI has not given a launch date or pricing for GPT-Synopsys. Hardware and FPGA engineers can borrow the eval setup: give the model the original RTL or a written spec, gate correctness with simulation, then estimate frequency with Yosys to compute throughput. If you use general models to rewrite RTL, always pair them with formal or equivalence checks before merging.

Action HubReady to adopt this in production?

Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.

Up NextLMSYS Launches the SGLang Consortium at the First SGLang Summit: NVIDIA, AMD, Intel, Google, Meta, Cloudflare, Modal and RadixArk Join as Founding Members to Help Shape the Open-Source Inference Engine Roadmap
Read Now→