It will be most beneficial where there is reuse, he believes, outlining how AI can be taught a design, and self-learning will allow it to adapt for a new design. As a large part of chip design is derivative, reuse helps the AI learning process, he pointed out.
AI in the design community is inevitable, he said but, just as in the first industrial revolution, some machines replaced some manual jobs, this can be a positive, argued Andersen.
“We would look back and say, who would want to do this very laborious task. . . especially when it comes to a physical task. . . . If you use those tools right, then you essentially elevate the task. And the same thing happens in chip design. If we if I provide a technology that automates a generation in our world of things like RTL or timing constraints or it automatically figures out that somebody misspelled something in the script or fixes errors, I think most people say “This is great because now I can be more productive”. The positives are that designers can spend more time on creative work. AI is good at automating tasks, but coming up with like brand new ideas, is still very much a human domain, he said.
“[AI] is replacing tasks, but it isn’t directly replacing engineers because the engineers can do more. And one more thing, it’s not that there is a million integrated circuits being created and now I need less people to do it. It’s quite the opposite. We don’t have enough people to to do essentially those chip designs. In fact, there’s many projections from various consulting companies that project like up to 30% work for a shortage. So from that perspective, I think everybody very much welcomes that because we can actually deal with the requirements that are being brought on to the chip industry.”
Synopsys will host SNUG Silicon 11-12 March 2026 at Santa Clara Convention Center, California.
