PodcastSupply ChainedSeptember 29, 2026
Adventures in Chip Design
Foundry capacity and system integration, not AI-EDA tooling, will determine which AI chip startups actually succeed.
- LLMs writing RTL: John reports LLMs are getting surprisingly good at one-shotting small chip design blocks in RTL (Register Transfer Level), despite little training data existing for it.
- Hallucination-like gaps persist: Models can't yet replicate the intuition of a veteran chip designer (e.g. OpenAI's Richard Ho) about "what might go wrong"—critical for verification.
The full card is a free trial away.
The full summary and the deep dive on this episode, plus a brief every morning on the shows and companies you follow.
Read free — start your trial14-day free trial
Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.
