Intel's Desmond Kirkpatrick demonstrates how LLMs can assist in real hardware design, not just software. Using the ROHD framework, AI works inside the design loop to develop test-driven, synthesizable hardware using Dart.
Hey, I'm the lead for that! I think it addresses a lot of the problems raised in these comments about Chisel and other generator frameworks. Happy to answer questions! https://intel.github.io/rohd-website/
Hardware generator frameworks allow you to use a full-powered modern software language to control generation of hardware. Think about why it may be easier to implement something in Python than C. Adding capabilities to VHDL/SystemVerilog is very difficult since many vendors and tools have to implement it, and if any of them don't support it the whole feature can't be used.
ROHD (https://intel.github.io/rohd-website/) is sort of similar to Chisel on the design side (Dart instead of Scala), plus a big focus on verification. It has explicit clock and reset (though nothing precludes automating connectivity of clocks and reset via your generation software).
You may find ROHD to be interesting, since it takes a very serious approach to verification and even has a simulator built-in!
https://intel.github.io/rohd-website/
Video demo: https://www.youtube.com/watch?v=SAPAi8Y4Z68 Blog post: https://intel.github.io/rohd-website/blog/ai-accelerated-agi...
The approach combines agile principles, built-in simulation, and AI-guided iteration.