Synopsys: Autonomous AI Agents to Tame Chip Design Complexity

Find the Latest SoC Solutions
for... Automotive... IoT... Security... Audio... Video

eetimes.com, Mar. 20, 2025 – 

At the annual Synopsys SNUG conference here in Santa Clara, Calif., the company’s CEO Sassine Ghazi spelled out the path to using AI agents to “tame” the increasing complexity of chip design.

In his keynote speech on Wednesday at the opening of the 35th year of the SNUG conference, Ghazi talked about a path analogous to the automotive industry’s transition from advanced driver-assistance systems (ADAS) to full autonomy, with an agentic AI development framework that envisions progression from foundational AI capabilities to fully autonomous multi-agent systems capable of high-level decision making and actions.

He emphasized the need to “re-engineer” engineering in the era of pervasive intelligence, and how this meant it was a special time to be an engineer to create the innovation required for AI.

Highlighting Nvidia CEO Jensen Huang’s comments that he would need a million Synopsys engineers to address design complexity, Ghazi implied this could be addressed with AI agents working in partnership with human engineers.

The challenge, Ghazi said, is that complexity is impacting every industry and hence engineering teams face exponential challenges in delivering future products. So, agents—and ultimately autonomous collaborative AI agents—will be the answer. This was re-enforced in his fireside chat with Arm CEO Rene Haas later in the day, when they talked about getting from ideation to tapeout in less and less time—especially with expectations of typical design cycles needing to get shorter while chips get even more complex. This expectation of faster product development cycles will accelerate further as Nvidia’s Huang also talked the day before in his keynote about a yearly cadence of new product introductions.

“Synopsys’ new generative AI-powered assistive and creative capabilities are unlocking additional levels of efficiency for customers, improving their designs and accelerating their time to market,” Ghazi said. “Still, when it comes to AI-driven EDA, we’re just scratching the surface of what’s possible. We see a future, paradigm-shifting opportunity with agentic AI, which will allow R&D teams to focus on important architecture and design decisions while tasking agentic engineers with implementation details.”

Click here to read more...

 Back