OpenAI and Synopsys Launch GPT-Synopsys AI for Chip Design

OpenAI and Synopsys have signed a multi-year partnership to build GPT-Synopsys, a specialized AI model trained to operate chip-design software the way an expert engineer does. The deal, first reported by Reuters on September 30 and confirmed in a formal announcement on October 2, pairs OpenAI's frontier AI models with Synopsys' industry-leading electronic design automation (EDA) tools. The goal: hand off chip-design objectives to AI agents that can run the tools, interpret results, and iterate toward verified outcomes — potentially cutting weeks or months off the design process.

According to the companies, GPT-Synopsys goes beyond today's AI coding assistants. Current agentic AI tools can connect general-purpose models to EDA software, but this partnership aims to make frontier models genuine experts in using those tools: learning to run them like seasoned engineers, reading their outputs, and iteratively optimizing designs. Engineers will delegate objectives — from optimizing power, performance, and area (known as PPA) to closing timing and verification issues — and agents will do the heavy lifting before a human reviews the final result.

What GPT-Synopsys actually does

In a joint statement, the companies described GPT-Synopsys as an AI-native approach to chip design. The model will reason about chip design and verification, then directly operate Synopsys' software: running simulations, implementing changes, and iterating toward verified outcomes for engineer review. It will run on OpenAI-hosted infrastructure, interoperate with customers' own agent systems, and integrate deeply with Synopsys.ai and the Synopsys Autopilot agentic AI platform. Early technology engagements with leading semiconductor customers are already underway, according to the October 2 announcement.

One important detail: the AI won't get the final word. Synopsys CEO Sassine Ghazi told Reuters that the model's work will still be double-checked by Synopsys tools that use traditional computing techniques to verify whether a chip will actually work. The model needs these guardrails "in order to check the physics," Ghazi said, adding that validating designs with the highest level of fidelity — what the industry calls sign-off, or ground truth — remains essential. OpenAI co-founder Greg Brockman said the goal was to "shave off weeks, months from the design process and to bring more chips to the world."

Why investors and engineers are paying attention

The money side of the deal is notable. The agreement includes a revenue-sharing arrangement and go-to-market collaboration to make GPT-Synopsys available to customers worldwide. OpenAI will pay Synopsys a training subscription fee while the model learns to use its tools, and when customers deploy the product, the two companies will share revenue based on how much the model improves a chip's design. Ghazi told Reuters the deal was deliberately structured so it adds revenue rather than replacing Synopsys' existing business, describing the arrangement as pure upside from delivering more value to customers.

Wall Street liked what it heard. Synopsys shares jumped as much as 13% after the GPT-Synopsys announcements, and the company raised its fiscal 2027 revenue outlook to roughly $11 billion to $11.2 billion — about 15% growth, well above the 11.19% analysts had expected. The same week also brought a separate $1 billion multi-year agreement between Synopsys and Amazon to develop custom chips, AI products, and cloud infrastructure, underscoring how aggressively big tech is moving upstream from chip factories into the software that designs the chips.

Why does this matter for anyone who isn't a chip engineer? Modern life runs on semiconductors, and designing them is one of the most complex, time-consuming jobs in tech — a single chip design can involve billions of transistors arranged on a sliver of silicon. If GPT-Synopsys can safely compress that timeline, it could speed up everything from AI accelerators to the chips in your phone and laptop. There's also a security angle built into the announcement: the joint offering promises enterprise-grade governance, with customer design data encrypted, never used to train the model, and controlled through configurable retention and audit tools.

Zoom out, and the partnership fits a bigger pattern in AI news: the technology is now being used to build itself. In the statement, OpenAI's Brockman described the effort as using the company's most advanced technology to improve the systems that power AI, bringing that work to chip design so engineers can explore more designs and reach a working chip faster. Whether GPT-Synopsys delivers on that promise will depend on how well agentic AI handles the unforgiving physics of real silicon — but for an industry that lives on timelines measured in months, even modest acceleration is a big deal. Read more on the GenZ NewZ homepage for continuing coverage.