I work in semiconductors, and the very best “AI” has to offer in this space is just searching documentation. It is fundamentally useless at anything silicon-design related, much to the dismay of EDA tool vendors who would really like you to buy more licenses for shit you don’t need.
These articles tend to overstate “using AI to design” technical systems when conceptually they’re really fuzzing the performance characteristics of different variations on an algorithm against a reference implementation, and validating its correctness.
I’m coming from software so I’m not sure how much simulation/verification can be done automatically in chip design, but I’d bet money this sort of process is what they’re talking about, similar to the alpha evolve paper from a bit ago. Rather than going “hey chatgpt how do I design this chip??”
Imagine using generative AI to design a chip, paying millions of dollars for the tapeout, only to find that the chip doesnt work. The AI won’t be held accountable for mistakes like that, but the engineer that used it sure will be.
I was about to say something like "as if management would let you send a design for tapeout without extensive verification and simulation but management that allows you to use genial for this would be exactly the type to not do that 😭
Sure grandpa, let’s get you to bed.
I work in semiconductors, and the very best “AI” has to offer in this space is just searching documentation. It is fundamentally useless at anything silicon-design related, much to the dismay of EDA tool vendors who would really like you to buy more licenses for shit you don’t need.
These articles tend to overstate “using AI to design” technical systems when conceptually they’re really fuzzing the performance characteristics of different variations on an algorithm against a reference implementation, and validating its correctness.
I’m coming from software so I’m not sure how much simulation/verification can be done automatically in chip design, but I’d bet money this sort of process is what they’re talking about, similar to the alpha evolve paper from a bit ago. Rather than going “hey chatgpt how do I design this chip??”
Imagine using generative AI to design a chip, paying millions of dollars for the tapeout, only to find that the chip doesnt work. The AI won’t be held accountable for mistakes like that, but the engineer that used it sure will be.
I was about to say something like "as if management would let you send a design for tapeout without extensive verification and simulation but management that allows you to use genial for this would be exactly the type to not do that 😭