San Francisco/Sunnyvale: OpenAI and semiconductor engineering company Synopsys have entered into a multi-year strategic partnership to develop GPT-Synopsys, a specialised artificial intelligence model designed to assist engineers with complex semiconductor design and verification tasks.

The partnership brings OpenAI’s advanced AI capabilities together with Synopsys’ electronic design automation (EDA) software and chip-engineering expertise. The companies aim to use AI agents to accelerate the development of increasingly complex computer chips.

AI Model Will Work Directly With Chip-Design Tools

GPT-Synopsys is being developed to reason about semiconductor design and directly operate Synopsys’ EDA tools.

Instead of simply providing suggestions to engineers, AI agents will be able to receive design objectives, run specialised engineering tools, interpret their results and repeatedly modify designs. Areas include power, performance and area optimisation, timing and verification.

This approach could allow engineers to evaluate a much larger number of design possibilities in less time.

Potential to Reduce Chip Development Time

Modern semiconductor design involves extremely complicated processes, including the arrangement of billions of transistors and repeated testing of different design configurations.

OpenAI co-founder Greg Brockman said the technology is intended to help engineers reduce development time by potentially weeks or months and accelerate the delivery of new chips.

For the semiconductor industry, even relatively small reductions in development time can have significant commercial value because chip development requires extensive engineering resources and lengthy verification cycles.

OpenAI to License Synopsys Software

As part of the agreement, OpenAI will license Synopsys’ EDA tools to develop and train the specialised model.

The companies will also collaborate on research and development, commercialization and customer engagement. Their agreement includes a shared-revenue framework, meaning both companies are positioned to benefit when customers use the resulting technology.

Early technology engagements with semiconductor customers are already underway.

Human and Software Verification Remain Important

Despite the increasing role of AI, the companies do not intend to eliminate conventional verification.

GPT-Synopsys’ proposed workflow will use Synopsys’ established engineering tools to check AI-generated design changes. This is particularly important because a semiconductor design must satisfy physical and electrical requirements before it can move toward manufacturing.

Synopsys CEO Sassine Ghazi has emphasised that high-fidelity verification remains essential for ensuring that AI-assisted designs can actually work in physical silicon.

Enterprise Security Built Into the Service

The companies say GPT-Synopsys will include enterprise-level security, governance and access controls.

Customer-specific design information is expected to be encrypted during transmission and while stored. The companies also state that customer data will not be used to train the model, while configurable retention, auditing and permission controls will be available.

These safeguards could be particularly important because semiconductor companies typically work with highly valuable intellectual property and confidential chip architectures.

AI Could Change Semiconductor Engineering

The partnership represents a broader shift in how artificial intelligence is being applied to technology development.

Rather than using AI only for software coding, customer support or content generation, companies are increasingly applying advanced models to highly specialised engineering tasks.

If GPT-Synopsys performs as intended, semiconductor engineers could use AI agents as collaborative engineering systems capable of exploring design alternatives, identifying optimisation opportunities and executing repetitive design workflows.

The technology could ultimately help chip companies shorten development cycles while allowing engineers to concentrate more heavily on high-level architecture, validation and strategic decisions.

A Strategic Partnership for the AI Hardware Era

The relationship between OpenAI and Synopsys also reflects a fundamental feature of the current AI industry: better AI requires increasingly sophisticated computing hardware.

As AI models become more capable, demand for advanced processors and specialised accelerators is increasing. Faster and more efficient chip design could therefore help accelerate the development of the very hardware needed to power future AI systems.

The partnership could mark an important step toward an AI-driven semiconductor design process in which frontier AI models and specialised engineering software work together to develop next-generation chips.