OpenAI and Synopsys are jointly building an AI model for chip design

OpenAI and Synopsys announced a strategic partnership on 30 September 2026 to build GPT-Synopsys, an AI model capable of autonomously executing chip designs using Synopsys' EDA tools. The model is intended to function like an experienced chip engineer: operating the software, optimising designs and making design decisions without continuous human intervention.

Major semiconductor companies are already testing the model in early programmes. Synopsys shares (NASDAQ: SNPS) rose seven percent following the announcement. Synopsys simultaneously raised its revenue growth outlook for fiscal 2027 to fifteen percent, well above the analyst estimate of 11.19 percent.

The partnership connects two parties with complementary positions: OpenAI brings its models and training infrastructure, while Synopsys contributes its dominant position in EDA software and years of chip design data.

What GPT-Synopsys does and how it is built

In traditional chip design, engineers configure EDA tools manually, a time-consuming process that requires deep domain knowledge. GPT-Synopsys is designed to take over that work: the model independently navigates Synopsys' tools, interprets design requirements and optimises outcomes.

OpenAI licenses Synopsys' EDA tools to train the model. OpenAI trains the model using Synopsys' EDA tools and hosts the infrastructure on which GPT-Synopsys runs. The model is deeply integrated with Synopsys.ai and the Synopsys Autopilot platform, an agentic AI platform whose details Synopsys disclosed on 28 September 2026 and which is set to become available at the end of 2026.

Customer data is not used for training. Data is stored encrypted and managed with security controls, a point Synopsys explicitly emphasises to its customers in the semiconductor industry, where intellectual property is highly sensitive.

Commercial structure and financial arrangements

The deal operates on a shared revenue model. OpenAI pays Synopsys a subscription fee for the use of the EDA tools during training. Thereafter, both companies share revenues based on the model's performance, specifically, the degree to which GPT-Synopsys improves customers' chip designs partly determines how revenues are divided.

Synopsys has a market capitalisation of approximately $82.79 billion, making it one of the larger players in the EDA market. The raised revenue growth forecast of fifteen percent for fiscal 2027 is partly linked to the expected contribution of this partnership and the broader Autopilot platform, for which more than fifty active engagements are already underway with customers working with AgentEngineer solutions.

Alongside the OpenAI deal, Synopsys also announced an IP agreement with Amazon, part of its Application-Optimized IP model (AOIP). This points to a broader strategy to monetise existing IP portfolios through cloud players and AI companies.

OpenAI's own chip interests are a factor

The partnership serves a dual purpose for OpenAI. The company has been working on its own chip strategy for some time to reduce its dependence on external suppliers such as Nvidia. Access to Synopsys' EDA tools and expertise ties directly into that: any company that wants to design its own chips needs precisely these tools.

Greg Brockman, president and co-founder of OpenAI, and Sassine Ghazi, president and CEO of Synopsys, signed the agreement. Concrete statements from either executive regarding their strategic rationale cannot be verified based on the available source material.

GPT-Synopsys runs on infrastructure managed by OpenAI, meaning chip companies that wish to use the model become dependent on OpenAI's cloud environment. This is a relevant consideration for companies that, for security or sovereignty reasons, prefer to keep data on-premise or within a European environment.

Implications for the European semiconductor and AI landscape

For European chip companies and fabless designers, including parties that benefit from the European Chips Act ecosystem, this is a signal that the EDA market is changing rapidly. Large semiconductor companies are already testing GPT-Synopsys, which means the bar for engineering capability and design speed is set to rise in the near term.

At the same time, the deal illustrates a pattern that is increasingly visible: American AI labs are forming exclusive or semi-exclusive partnerships with sector-specific data holders, concentrating access to specialised AI models within a small group of platforms. For European policymakers working on AI regulation and technological autonomy, this is a development worth monitoring, particularly in a sector like semiconductors, which is regarded as strategically critical.