Training Data How Ricursive Intelligence’s Founders are Using AI to Shape The Future of Chip Design
90 snips
Jan 14, 2026 
Azalia Mirhoseini and Anna Goldie, co-founders of Recursive Intelligence, revolutionized chip design at Google with AlphaChip, drastically speeding up the process. They discuss how chip design bottlenecks hinder AI's progress and their vision for 'designless' custom silicon, making it accessible for all companies. The duo shares insights into using AI for advanced placements, novel chip shapes, and recursive self-improvement, where AI enhances its own designs. Their optimism for AGI and a 'Cambrian explosion' of custom silicon applications paints an exciting future for technology.
AI Snips
Chapters
Transcript
Episode notes
Chip Design Is The Compute Bottleneck
- Chip design is the current compute bottleneck slowing AI progress because chip cycles lag model innovation.
- Co-designing models and hardware would unlock much greater effective compute and faster AI scaling.
From Research To Tape-Out With TPU Team
- Azalia and Anna started with compiler-to-hardware mapping work before moving into chip placement at Google in 2018.
- They partnered closely with the TPU team and iterated until research moved into production and tape-out.
Floorplanning Is A Massive Combinatorial Problem
- Floorplanning places millions of chip nodes under tight PPA and physical constraints, making it a massive combinatorial optimization.
- Evaluating metrics is costly and complex, which limits how quickly designers can iterate.