LiquidJet is a 3D ultra short-loop coldplate that features unique multi-stage cooling architecture and is customizable to any GPU power map.​

LiquidJet Performance on NVIDIA GB300 Compared to Skived Coldplate:​

• 75% higher heat transfer efficiency
• 6 – 12 ºC cooler GPU
• 10 – 25% higher AI Tokens per watt​​
• Seamless drop-in upgrade
  

Future-proofed for what’s next, LiquidJet is engineered to scale with next-gen GPUs like NVIDIA Rubin, Rubin Ultra, Feynman (>4,000W) and beyond, as well as for custom hyperscaler ASIC designs. As AI workloads explode, Frore Systems’ LiquidJet architecture is built to keep up.

Performance on the Blackwell Ultra can be seen below:

‍

LiquidJet’s revolutionary direct-to-chip liquid cooling 3D coldplate solution is designed to meet the escalating demands of AI Data Centers. LiquidJet unlocks higher performance from the world’s most powerful GPUs—starting with NVIDIA Blackwell Ultra—delivering significant reductions in data center total cost of ownership (TCO) and improving power usage effectiveness (PUE).

‍

Hotspot Cooling with LiquidJet 

LiquidJet’s innovative multi-stage architecture enables the entire flow to be concentrated to first cool the hotspots, before cooling the other areas of the GPU die.  

LiquidJet supports hotspot cooling up to 670 W/cm², far beyond the requirements of future GPUs: Rubin, Rubin Ultra, Feynman, and custom ASICs.

‍

‍

Customizable Cooling

Legacy coldplates rely on 2D flow through skived microchannels, limiting their ability to adapt to new chip layouts and increasingly non-uniform high power density requirements.

Frore Systems takes a different approach to LiquidJet coldplate manufacturing, adapting semiconductor processes to metal wafers, fabricating multi-stage cooling architecture with innovative 3D ultra short-loops that are designed precisely to the power maps of modern GPUs.  ​

‍

Customizable to match the exact power density of your GPU.

The result is LiquidJet, a coldplate that evolves as fast as the chips it cools. 

‍

With these advances, data centers can achieve:

• Cooler GPUs
• More AI Factory Tokens/Watt​
• Lower total cost of ownership (TCO) 
• Improved power usage effectiveness (PUE)
• Higher reliability

LiquidJet - Redefining AI Data Center Liquid Cooling.