OpenAI and Synopsys Partner to Build GPT-Synopsys for Autonomous Chip Design

OpenAI and Synopsys Form Strategic Partnership for AI Chip Design

OpenAI and Synopsys announced a multi-year agreement on September 30 to jointly develop GPT-Synopsys, a specialized AI model for semiconductor design that combines OpenAI's frontier AI capabilities with Synopsys's electronic design automation tools and domain expertise1

1

2

2

. Under this revenue-sharing partnership, OpenAI will license Synopsys's EDA tools to develop the model, which will run on OpenAI-hosted infrastructure4

4

. The announcement came during Synopsys's investor summit, where CEO Sassine Ghazi revealed the company expects 15% revenue growth for fiscal 2027, surpassing analyst estimates of 11.19%2

2

. Synopsys shares rose as much as 7% following the partnership announcement5

5

.

How GPT-Synopsys Will Transform Autonomous Chip Design

GPT-Synopsys represents a fundamental shift in how engineers approach semiconductor development. Unlike general-purpose AI models simply connected to engineering software, this specialized AI model for semiconductor design is being trained to become an expert user of electronic design automation tools themselves3

3

. Engineers will delegate objectives involving power-performance-area optimization, timing, and verification closure to AI agents that run workflows, interpret results, implement changes, and iterate toward verified outcomes1

1

. OpenAI co-founder Greg Brockman explained the model will help engineers work through design trade-offs and optimizations to "shave off weeks, months from the design process and bring more chips to the world"2

2

. The system handles tasks from circuit descriptions in code-like language to determining transistor layout across billions of components on silicon5

5

.

Revenue Structure and Business Model Innovation

The partnership establishes a novel financial arrangement where OpenAI pays Synopsys a training subscription fee for learning to use Synopsys tools2

2

. When semiconductor customers use the product, both companies will share revenue based on how well the AI model for chip design work improves chip design outcomes5

5

. Sassine Ghazi emphasized the agreement was structured to avoid cannibalizing existing business, stating it "will be an upside to our business given we're delivering more value to the customer"2

2

. The companies will jointly offer bundled compute, model, and software licenses needed to use GPT-Synopsys, alongside joint marketing initiatives to reach customers globally4

4

.

Integration with Synopsys.ai and Autopilot Platform

GPT-Synopsys will integrate with Synopsys.ai and the Synopsys Autopilot platform, the company's agentic engineering system announced just two days before the OpenAI deal1

1

. The Autopilot platform includes AgentEngineer, a portfolio of seven long-horizon agents covering verification, implementation, analog design, manufacturing, and other specialized tasks1

1

. Synopsys reported deployments demonstrating up to 50x faster verification closure, 20% higher coverage, and a 30% productivity boost3

3

. The system is designed to work with customers' existing agent harnesses, with Autopilot providing services such as memory and governance1

1

.

Verification Standards and Safety Guardrails

Despite the autonomous capabilities, GPT-Synopsys maintains rigorous verification standards. Synopsys's EDA engines will perform the actual calculations and checks, with the model interpreting results and agent software managing execution1

1

. Ghazi stressed that the model's work will be double-checked by Synopsys tools using traditional computing techniques to verify whether a chip will function correctly2

2

. "The model needs these guardrails in order to check the physics," Ghazi explained. "The need for validating with the highest level of fidelity, what we call sign-off or ground truth, is essential"5

5

. This approach ensures manufacturing compliance before tapeout for fabrication1

1

.

Data Privacy and Security Measures

Addressing concerns about proprietary design data, Synopsys confirmed that customer-specific chip-design data will not be used to train the model3

3

. Data privacy and security protocols include encryption both at rest and in transit, with customers maintaining configurable controls covering retention, permissions, and auditing3

3

4

4

. These measures address critical concerns for semiconductor customers who handle highly sensitive intellectual property throughout the design process.

OpenAI's Expanding Chip Development Footprint

This partnership builds on OpenAI's growing involvement in chip development. Earlier in June, the company unveiled Jalapeño, its first custom inference accelerator developed with Broadcom1

1

3

3

. The chip went from initial design to manufacturing tape-out in nine months, with OpenAI's models used to accelerate the chip design process and optimization3

3

. Greg Brockman stated, "We're using our most advanced technology to improve the systems that power AI. By helping them build better chips, we can build better AI and bring it to more people"3

3

. GPT-Synopsys effectively transforms that internal experiment into a commercial product available to the broader semiconductor industry.

Market Context and Competitive Landscape

The move toward agentic AI in chip design extends beyond Synopsys. Competitors Cadence and Siemens EDA are also developing autonomous design agents1

1

. Synopsys previously launched DSO.ai in March 2020, a reinforcement learning tool for power-performance-area optimization, followed by Synopsys.ai Copilot in November 2023 through collaboration with Microsoft's Azure OpenAI service1

1

. Early technology engagements with leading semiconductor customers are already underway, though companies did not name specific partners3

3

4

4

. Synopsys reported more than 50 engagements with availability planned for late 20261

1

. The immediate audience includes professional semiconductor-design teams and companies developing custom silicon1

1

.