OpenAI has announced a major partnership with software developer Synopsys to create a specialized artificial intelligence model called GPT-Synopsys. This project aims to completely revolutionize semiconductor design by eliminating routine operations and minimizing the human factor in the early stages of microchip creation. As of October 2026, the semiconductor industry is facing a colossal deficit of time and resources, making such breakthroughs vital for maintaining the pace of technological progress.
The New Role of Artificial Intelligence in Microelectronics
The developers emphasize that GPT-Synopsys is not created as a regular auxiliary chatbot integrated into software, but as a fully autonomous expert user. Engineers only set general target parameters such as power consumption, overall performance, die area, timings, and verification requirements (PPA metrics). Afterward, AI agents take over the management of specialized EDA tools, independently analyze source data, and make adjustments to the highly complex architecture of future semiconductors.
Autonomy and the Iterative Verification Process
The chip creation process using GPT-Synopsys is built on continuous iterations. The system independently tests every step until it obtains a fully verified design, which is then submitted for final expert review by human engineers. As Synopsys CEO Sassine Ghazi notes, the future of the semiconductor industry requires a radical acceleration of development without the slightest loss of quality or the risk of critical errors during the initial silicon manufacturing run. The new model operates on OpenAI's secure infrastructure and is deeply integrated with the Synopsys.ai platform and Synopsys Autopilot agent environment.
Corporate Data Security and Confidentiality
Special attention in the project is paid to information security and the protection of corporate clients' intellectual property. Synopsys management officially states that confidential customer data and their proprietary developments will under no circumstances be used for further training of the GPT-Synopsys model. All information is encrypted both during transmission across networks and in dedicated secure storage, while clients receive advanced access control and detailed audit tools.
