OpenAI and Synopsys Partner to Build AI-Native Chip Design Model

The collaboration shifts chip design from manual interaction toward agentic workflows. Instead of traditional human-led processes, GPT-Synopsys will function as an expert engineer, capable of operating Synopsys EDA software, interpreting complex outputs, and iteratively refining designs. Engineers will delegate specific objectives—such as timing closure and verification—to the system, which will handle the underlying tool execution and implementation tasks.

Sassine Ghazi, CEO of Synopsys, emphasized that the goal is to drive design acceleration without sacrificing the rigor required for manufacturing. For OpenAI, the partnership represents a strategic move to improve the foundational hardware that sustains its own AI infrastructure. Greg Brockman, president and co-founder of OpenAI, noted that by assisting engineers in building superior chips, the companies hope to scale the accessibility and performance of future AI systems.

To address enterprise concerns, the companies confirmed that GPT-Synopsys will operate on OpenAI-hosted infrastructure with strict security protocols. Customer design data will remain protected, encrypted, and excluded from the model’s training sets. The joint offering is already undergoing early engagements with select semiconductor customers, aiming to provide a bundled solution that includes compute, model access, and software licensing.