Frore Systems lands $143M to scale liquid cooling for AI data centers
Frore Systems, a San Jose-based startup building liquid cooling technology for high-performance AI hardware, has raised $143 million in fresh funding at a reported $1.64 billion valuation, according to Bloomberg. The round was led by MVP Ventures and included participation from Fidelity Management & Research, Top Tier, Mayfield Fund, and Qualcomm Ventures.
The company said it will use the new capital to expand manufacturing capacity as demand grows for cooling systems that can keep increasingly power-hungry AI chips operating efficiently.
From fanless devices to data-center infrastructure
Founded about eight years ago by Madhavapeddy and Suryaprakash Ganti, Frore Systems initially focused on compact cooling for smartphones and other electronics that do not rely on traditional fans. As AI computing accelerated—particularly systems built around NVIDIA GPUs—the company shifted toward liquid cooling designed for server-class hardware.
LiquidJet Nexus targets higher heat loads
Frore’s flagship product, the LiquidJet Nexus tray, circulates coolant through three-dimensional channels engineered for specific chip types. The company contrasts this approach with conventional liquid cooling built around flat cold plates, arguing its channel design improves heat removal and enables coolant to cycle through the system multiple times.
Frore also says its materials and form factor can make cooling assemblies thinner and lighter, allowing more components to be stacked within servers—potentially improving compute density while reducing electricity use, water consumption, and cooling-related costs.
Manufacturing expansion and customer demand
Frore currently manufactures in Taiwan and plans to add capacity and additional production locations. The company counts large cloud providers, governments building national compute networks, and AI server hardware makers among its customers, though it has not publicly disclosed client names.
The pivot toward data-center liquid cooling was influenced by conversations with NVIDIA CEO Jensen Huang, who encouraged adapting the technology to AI infrastructure, the company has said.










