Nicolas Sauvage, head of TDK Ventures (managing $500M across four funds), outlines a thesis that major AI infrastructure bets become obvious only after four years. The article profiles Groq, a $6.9B inference-focused AI chip startup founded by Jonathan Ross (a Google TPU engineer), which Sauvage backed in 2020 before the generative AI boom. Sauvage argues that inference demand will continue compounding as AI agents proliferate, making unsexy but essential infrastructure the better long-term play.
Infrastructure
Nicolas Sauvage is betting on the boring parts of AI
Sauvage's backing of Groq exemplifies the contrarian play that unglamorous but essential AI inference infrastructure becomes the durable winner as agentic AI compounds demand.
Monday, May 4, 2026 12:00 PM UTC2 MIN READSOURCE: TechCrunchBY sys://pipeline
Tags
infrastructure