Ricursive Intelligence Bags $300 Million in Funding, Hits $4 Billion in Valuation Just 2 Months After Launch

Ricursive Intelligence, a frontier AI Lab focused on building self-improving systems, starting with chip design, led by Anna Goldie, Azalia Mirhoseini, and others, has secured $300 million in funding led by Lightspeed with participation from DST Global, Nvidia’s venture capital arm NVentures, Felicis Ventures, 49 Palms Ventures, and Radical AI.

Ricursive aims to accelerate AI chip innovation by using AI systems to design silicon substrates and iteratively improve hardware toward AGI. The company raised a Series A two months after launching from seed funding and has raised $335 million in total. Founded by former Google researchers, its team previously developed AlphaChip, a reinforcement learning system used in multiple generations of Google TPU chips.

Traditional chip design cycles take 2–3 years, creating a lag where hardware trails algorithmic progress. AI–chip co-evolution compresses design timelines to weeks, forming a recursive loop where AI designs better chips, those chips enable stronger AI, and the cycle repeats. This shift turns chip design into an accelerant for compute and intelligence scaling, with a roadmap toward autonomous end-to-end semiconductor design and continuous AI–silicon co-improvement for frontier AI infrastructure.