Agentic AI, RISC-V and the Future of EDA
Amit Goldie, Senior Director of Product Engineering at Synopsys, brings extensive experience in semiconductor verification technology, electronic design automation (EDA), and product engineering. He oversees engineering for verification products spanning static verification, low-power design, constraints, and simulation, supporting a technology portfolio valued at over $500 million. His expertise also covers RISC-V ecosystem partnerships, cloud-based verification solutions, power optimization, and machine learning applications for improving product debuggability. With experience in global team leadership, enterprise customer engagement, and go-to-market strategy, alongside authorship of more than 30 technical papers, he brings both technical depth and commercial perspective to the evolving semiconductor landscape. Drawing on this experience, Amit explores how agentic AI, RISC-V architectures, and cloud-based EDA are reshaping semiconductor design and verification. The article examines how AI-driven workflows can automate routine verification and debugging tasks, how customizable RISC-V architectures increase the importance of rigorous verification, and how cloud infrastructure enables scalable simulation and faster development cycles. For investors, these developments offer a framework for evaluating opportunities across EDA software, semiconductor design infrastructure, and next-generation verification technologies. The central question is which companies can translate these technological advances into measurable productivity gains, shorter development timelines, and sustainable competitive advantages.
Q1. As EDA moves from AI-assisted tools toward agentic workflows, where are you seeing the biggest productivity gains in chip design and verification?
The biggest improvements are happening in areas where engineers usually spend a lot of time juggling tools, making sense of results, and figuring out their next steps. With AI-assisted EDA, specific tasks like generating assertions, analyzing logs, and suggesting possible failure causes have already become much quicker. Agentic workflows take this to the next level by linking these tasks together in an ongoing, streamlined process.
Verification
Verification is a great example of where this approach shines. An agent can keep an eye on regressions, sort failures, link them to recent design changes, and suggest what to debug next. It can also spot coverage gaps, propose new tests or constraints, kick off the right jobs, and check if coverage has improved. Right now, engineers often handle all these steps by hand, jumping between different tools and data sources. Cutting down on that manual coordination can save hours—or even days—during a big regression cycle.
Debug
Debugging is another area where this technology can really make a difference. A verification engineer might be faced with thousands of failures—often caused by …
Create an account to continue reading
Create AccountAlready have an account? Sign in
Need an expert in this space?
Talk to an Industry Expert
Knowledge Ridge connects decision-makers with carefully vetted subject matter experts for one-on-one calls, research sprints, and advisory engagements — across 11 sectors and 163 sub-industries globally.
Comments
No comments yet. Be the first to comment!