Ricursive Intelligence Raises $335M for AI Chip Design
Ricursive Intelligence, an AI chip design startup founded by former Google Brain engineers, has secured $335 million in funding at a $4 billion valuation. The company reached this valuation just four months after its inception. The investment underscores the strategic importance of developing custom hardware for large-scale agentic AI.
- Co-founders Dr. Anna Goldie and Dr. Azalia Mirhoseini previously led the creation of AlphaChip at Google, an AI system using reinforcement learning that automated chip layout design for multiple generations of Google's Tensor Processing Units (TPUs). - The company's core technology uses a reinforcement learning agent that iteratively generates and scores chip layouts to improve their quality, framing the physical design of chips as a solvable AI problem. This aims to shorten design cycles that typically take large teams two to three years and hundreds of millions of dollars. - The total $335 million raised comprises a $35 million seed round led by Sequoia Capital and a $300 million Series A led by Lightspeed Venture Partners. NVentures, the venture capital arm of Nvidia, is also a strategic investor. - Ricursive's strategy is not to manufacture chips, but to provide AI-powered electronic design automation (EDA) tools to any company wanting to create custom silicon, positioning itself as a "designless" enabler for the semiconductor industry. - The push for custom hardware is driven by the demands of agentic AI, whose complex workflows require heterogeneous systems with specific optimizations for memory, compute, and bandwidth to run efficiently and cost-effectively. - The investment reflects a wider trend of high-value bets on AI-designed hardware, with competitors like Unconventional AI, founded by a former Intel executive, also attracting significant funding at a multi-billion dollar valuation.
