Ricursive Intelligence launched a new frontier AI lab focused on accelerating semiconductor design with advanced AI models, securing a $35 million seed round led by Sequoia Capital at a $750 million valuation. The Palo Alto-based startup was founded by Anna Goldie and Azalia Mirhoseini, the AI researchers behind Google’s AlphaChip system that applied reinforcement learning to chip floorplanning. Their new venture seeks to shorten multi-year chip development cycles and build a recursive framework in which AI systems design the chips that will train the next generation of AI models.
The company is building a full-stack AI platform to optimize every stage of chip development—layout, optimization, verification, and architecture exploration—creating a continuous feedback loop between model and silicon. Goldie said the goal is to drastically reduce design time and enable more frequent iterations, directly impacting the compute available for frontier-scale AI systems. Mirhoseini noted that scaling AI has historically required advances in chip design, and the team intends for AI-generated silicon to unlock new model capabilities.
The investment from Sequoia signals growing momentum behind AI-accelerated electronic design automation (EDA), an area seeing heightened interest from semiconductor companies, hyperscalers, and emerging design-automation startups. With increasing demand for custom accelerators, energy-efficient architectures, and AI-optimized memory and interconnect systems, Ricursive’s approach positions it to influence the next wave of advanced chip development.
• $35M seed round led by Sequoia Capital at $750M final valuation
• Founded by Anna Goldie & Azalia Mirhoseini, creators of AlphaChip
• Platform applies AI to optimize full semiconductor design cycle
• Focus on recursive AI–chip co-development to accelerate hardware roadmaps
• Targets custom silicon for future frontier AI systems and superintelligence-scale compute
“Chips are the fuel for progress in AI, and the multi-year chip design process is holding back the field,” said Dr. Anna Goldie, Founder and CEO. “By closing the loop between AI and hardware, we can accelerate progress toward artificial superintelligence and enable a Cambrian explosion of custom silicon.”
🌐 Analysis
Ricursive Intelligence enters a fast-moving segment where AI-assisted chip development is gaining traction among EDA vendors and hyperscalers pursuing custom silicon. Their recursive “AI designs silicon → silicon trains AI” model aligns with broader industry shifts toward rapid tape-out cycles, heterogeneous packaging, and domain-specific accelerators. The founders’ AlphaChip lineage provides technical credibility at a time when both NVIDIA and Google are pushing AI-optimized chip design internally.
Dr. Anna Goldie and Dr. Azalia Mirhoseini both previously served as senior research scientists at Google DeepMind and Google Research, where they co-created AlphaChip, the reinforcement-learning system that demonstrated AI could outperform human experts in chip floorplanning—a milestone that reshaped thinking across the semiconductor and EDA industries. Goldie, now CEO of Ricursive Intelligence, specializes in large-scale optimization and machine-learning systems applied to physical chip design. Mirhoseini, now CTO, is known for pioneering work in deep reinforcement learning, hardware-software co-design, and automated architecture search.
