Sygaldry Raises $139 Million to Build Quantum-Accelerated AI Servers for Data Centers

Sygaldry Technologies, Inc. has secured $139 million in total financing to develop quantum-accelerated AI servers designed to drastically reduce the energy consumption of large-scale machine learning. The funding includes a $105 million Series A led by Breakthrough Energy Ventures and a $34 million Seed round led by Initialized Capital. Based in Ann Arbor and San Francisco, the company aims to integrate quantum hardware directly into existing data centers to speed up training and inference for increasingly large models.

The startup addresses a critical bottleneck as global demand for AI is projected to require approximately 125 gigawatts of new power generation capacity by 2030. Sygaldry’s servers are designed to operate alongside classical silicon within the data center, providing targeted acceleration for intensive AI algorithms. CEO and co-founder Chad Rigetti stated that the goal is to enable a fundamentally more efficient way of converting megawatts into intelligence.

Unlike traditional quantum systems that require extreme cooling, Sygaldry’s architecture combines multiple qubit types with room-temperature photonics for high-speed data transfer. This hybrid approach aims to eliminate the massive auxiliary equipment costs that have historically hindered quantum adoption. The company is also developing quantum-native algorithms that plug directly into the tools currently used by AI researchers, such as standard developer frameworks. Joining Rigetti in the venture are co-founders Idalia Friedson, a former strategy executive at Rigetti Computing, and AI scientist Michael Keiser. The investment syndicate features a wide array of deep-tech and academic backers, including Y Combinator, the University of Michigan, and QDNL Participations. Carmichael Roberts of Breakthrough Energy Ventures noted that the industry needs a breakthrough in performance per watt to bend the cost curve during this critical period of AI expansion.