OpenAI and Synopsys Partner to Build GPT-Synopsys AI Chip Design Model
The multi-year collaboration pairs OpenAI frontier models with Synopsys EDA tools to automate complex semiconductor workflows and optimize silicon architectures.
Key takeaways
- OpenAI and Synopsys signed a multi-year strategic partnership to develop GPT-Synopsys, a specialized model trained to operate electronic design automation (EDA) tools.
- The model will autonomously run workflows spanning circuit descriptions, transistor placement, timing closure, and power, performance, and area (PPA) optimization under engineer oversight.
- OpenAI will pay a training subscription fee and host the model, while both companies will share revenue and market a bundled offering containing compute, licenses, and AI.
- Synopsys verified that all outputs will undergo sign-off checks using traditional verification tools, and customer data will remain encrypted and excluded from training.
OpenAI and electronic design automation provider Synopsys have entered a multi-year strategic partnership to develop GPT-Synopsys, a specialized artificial intelligence model built to operate semiconductor design software autonomously. Announced on September 30, 2026, the collaboration aims to embed OpenAI's frontier reasoning capabilities directly into Synopsys engineering tools to assist human engineers with designing, verifying, and optimizing complex microchips, according to a joint announcement distributed via PR Newswire.
Under the agreement, OpenAI is licensing Synopsys EDA tools to train the specialized model. Rather than serving as an external general-purpose assistant, GPT-Synopsys is being trained to act as an expert user of Synopsys software, navigating chip architecture workflows from code-level circuit descriptions down to placing billions of transistors on silicon.
Autonomous Workflows and Verification Guardrails
GPT-Synopsys is designed to allow semiconductor engineers to assign high-level design targets—such as optimizing power, performance, and area (PPA), timing closure, and functional verification—to autonomous AI agents. The model will run Synopsys software, analyze intermediate simulation outputs, implement design adjustments, and iterate toward verified targets before submitting results for engineer review, as reported by StreetInsider.
To ensure silicon manufacturability, Synopsys emphasized that the AI model's work will be validated with classical physical simulation tools. In an interview with Reuters reported by The Indian Express, Synopsys CEO Sassine Ghazi explained that automated outputs will be double-checked using traditional computing techniques to verify underlying physics, noting that validating designs with sign-off ground truth remains necessary for first-time-right silicon.
OpenAI co-founder and president Greg Brockman stated that the initiative aims to shorten development timelines by weeks or months, speeding up the production of advanced microchips that in turn train and run future AI systems.
Infrastructure, Security, and Commercial Model
The specialized model will run on OpenAI-hosted infrastructure and integrate directly with the Synopsys.ai suite and the Synopsys Autopilot agentic platform, according to StreetInsider. The joint commercial offering will be sold as a bundled service combining OpenAI compute, the GPT-Synopsys model, and required Synopsys software licenses.
Addressing security concerns in proprietary chip development, both companies stated that customer design data will not be used to train OpenAI models. All customer information will be encrypted at rest and in transit, supported by configurable access, audit, and retention controls, according to Investing.com Canada.
Market Reception and Financial Outlook
The business arrangement includes both development fees and shared recurring revenue. OpenAI will pay Synopsys a training subscription fee to enable the model to learn EDA tool operations. When semiconductor customers deploy the resulting service, Synopsys and OpenAI will share revenue based on the measurable value and efficiency gains delivered to the customer.
Following the partnership announcement at an investor summit, Synopsys forecast 15% revenue growth for fiscal 2027, exceeding consensus analyst estimates of 11.19% compiled by LSEG, and Synopsys shares rose as much as 7%, as reported by The Indian Express. Early technology engagements and customer trials with semiconductor manufacturers are already underway worldwide.
Frequently asked questions
What is GPT-Synopsys?
GPT-Synopsys is a specialized AI model developed jointly by OpenAI and Synopsys. It is trained to directly operate electronic design automation (EDA) software, automate chip design workflows, and optimize semiconductor performance, power, and area.
Will customer semiconductor designs be used to train OpenAI models?
No. Synopsys and OpenAI confirmed that customer design data will not be used for model training. Customer data will remain encrypted at rest and in transit, with enterprise audit and retention controls.
How does GPT-Synopsys ensure that generated chip designs are physically accurate?
All designs generated or modified by GPT-Synopsys are double-checked and verified using Synopsys traditional simulation and sign-off computing tools to validate physical constraints and electrical functionality before fabrication.
Sources
- OpenAI and Synopsys team up to build an AI model that designs chips like a seasoned engineerThe Decoder · Sep 30, 2026
- OpenAI and Synopsys Announce GPT-Synopsys: Frontier Intelligence to Revolutionize Chip DesignCision PR Newswire · Sep 30, 2026
- Synopsys, OpenAI strike deal to develop AI model for chip design workindianexpress.com · Oct 1, 2026
- OpenAI and Synopsys partner to develop chip design AI modelstreetinsider.com · Sep 30, 2026
- Synopsys, OpenAI partner on AI model for chip designca.investing.com · Sep 30, 2026
How this story was made: the newsroom picked it up from the-decoder.com, gathered the full text of the sources above, and drafted it with AI assistance. Every factual claim was then checked against those sources before publishing (33 claims checked). Illustrations marked as AI-generated are not photographs. Spotted an error? Tell us.
Published October 1, 2026 at 07:42 UTC
