The Pulse
OpenAI and Synopsys Join Forces on GPT-Synopsys Chip-Design Model
OpenAI and Synopsys announced a multi-year partnership to develop GPT-Synopsys, a model designed to operate Synopsys electronic design automation tools in semiconductor workflows. The companies plan to jointly offer the service worldwide, b
AI.info Team ·
OpenAI and Synopsys announced on September 30 a multi-year agreement to develop GPT-Synopsys, a specialized model designed to operate Synopsys electronic design automation (EDA) tools during semiconductor design workflows. The companies will collaborate on research and development and joint marketing, and plan to offer the service worldwide under a shared-revenue arrangement. OpenAI will license Synopsys EDA tools for the model’s development. The announcement gives no launch date or pricing.
From tool access to tool expertise
The proposed model is meant to do more than connect a general-purpose AI system to design software. Synopsys says GPT-Synopsys will learn to run its tools, interpret their output and iteratively optimize designs. The companies describe it as combining OpenAI’s frontier models with Synopsys’ EDA technology and chip-design expertise, so it can reason about design and verification and directly operate the tools.
Engineers could delegate objectives ranging from power, performance and area (PPA) optimization to timing and verification closure. According to the announcement, agents would run tools, interpret results, make changes and repeat the process toward outcomes for engineer review. The companies say the approach is intended to help engineers explore more design alternatives and work on more sophisticated designs.
Synopsys framed the partnership around accelerating chip design without compromising PPA or first-time-right silicon. OpenAI president and co-founder Greg Brockman described the work as applying the company’s technology to chip design:
“With Synopsys, we're bringing that work to chip design, helping engineers explore more designs and get to a working chip faster.”
Greg Brockman, president and co-founder of OpenAI
OpenAI hosts the service; Synopsys supplies the tools
GPT-Synopsys is planned to run on OpenAI-hosted infrastructure and integrate with Synopsys.ai and the company’s Autopilot agent platform. Synopsys says it is also designed to interoperate with customers’ existing agent-harness systems. The planned joint service offering will bundle compute, the model and software licenses.
Early technology engagements with semiconductor customers are underway, Synopsys said, describing them as engagements with leading customers without naming participants. The announcement presents worldwide customer availability as a goal of the partnership, not as a product already generally available. It does not specify a schedule.
Design data protections and unanswered questions
Synopsys says customer design data will not be used to train the model and will be encrypted both in transit and at rest. The company also says customers will be able to manage configurable retention, audit and permission controls, alongside enterprise security, governance and access controls. These are stated protections for customer-specific design data; the announcement does not describe how they will be implemented in individual deployments.
The announcement does not identify the model’s technical architecture or the data used to specialize it. It provides no examples of completed chip-design tasks, named customer deployments, independent performance results or quantified time savings. Nor does it explain how design quality will be measured. Synopsys identifies development, capabilities, timing, availability and anticipated benefits as forward-looking statements, so the announced service and its expected outcomes remain plans rather than demonstrated results.