Frore Systems has introduced its latest cooling solution, LiquidJet Nexus, targeting next-generation AI server platforms such as Nvidia’s Kyber systems.
The new design builds on the company’s LiquidJet technology, using a direct-to-chip liquid cooling approach with a 3D jet-channel structure aimed at handling the growing thermal demands of modern AI GPUs. The company says the system is designed for compact half-U compute trays used in dense data center deployments.
According to Frore, LiquidJet Nexus improves heat transfer efficiency by up to 75 percent while keeping components several degrees cooler. The design also removes the need for traditional cooling hardware such as hoses, connectors, and manifolds, simplifying integration in high-density server environments. The cooling system relies on semiconductor-style manufacturing adapted to metal, enabling precise 3D microstructures that can be tuned to match GPU hotspot layouts. This allows higher heat removal efficiency and supports increasing power densities seen in modern and future AI accelerators.
LiquidJet Nexus is positioned as part of a broader “AI thermal stack,” with Frore aiming to address cooling as a system-level challenge rather than a standalone component. The approach is designed to improve overall data center efficiency, including power usage effectiveness and total cost of ownership.
Frore says the new solution is built to scale with upcoming GPU architectures, including multi-kilowatt designs expected in future AI platforms, where traditional cooling methods are starting to reach their limits. Additional details on availability and deployment timelines have not been announced, but the company is positioning LiquidJet Nexus as a drop-in solution for next-generation AI infrastructure.
Frore Systems bets on the AI thermal stack
Frore Systems is also highlighting a broader shift in AI infrastructure, where cooling is becoming a core part of system design rather than an afterthought. The company recently reached a $1.64 billion valuation and has achieved Unicorn status following the close of a $143 million Series D financing round. The Series D round was led by MVP Ventures and included participation from investors Fidelity Management & Research Company, Top Tier, Mayfield Fund, Clear Ventures, Addition, Qualcomm Ventures, StepStone Group, and Alumni Ventures.
The AI thermal stack spans everything from heat extraction at the chip level to heat rejection at the data center level, and is increasingly seen as a limiting factor for scaling AI workloads. Frore claims its LiquidJet and related technologies can improve efficiency, reduce power usage, and enable higher compute density across data centers, edge systems, and consumer AI devices. Frore Systems is targeting a broad range of AI platforms, including AI Data Centers, Industrial Edge AI Gateways, and Consumer AI Devices.
