Liquid Cooling is a thermal management approach that uses a liquid medium—typically water or a dielectric fluid—to absorb and transport heat away from IT components. Because liquids have far higher thermal conductivity and heat capacity than air, liquid cooling can handle significantly greater heat loads in a more compact and energy-efficient manner.
For a broader overview of liquid cooling technologies, applications, and deployment considerations, see ATTOM’s Liquid Cooling for Data Centers guide.
Liquid Cooling Types
- Direct-to-Chip (Cold Plate / D2C): Coolant flows through cold plates mounted directly on high-power components such as CPUs and GPUs, removing heat at the source with high precision.
- Rear Door Heat Exchanger (RDHx): A liquid-cooled door is attached to the rear of the rack, capturing hot exhaust air and transferring the heat into a coolant loop without requiring changes to the servers themselves.
- Immersion Cooling: Servers or components are fully submerged in a dielectric fluid; heat is transferred directly from the hardware surface into the liquid (available in single-phase and two-phase variants).
Liquid cooling has become essential for modern high-density environments. It enables rack power densities well beyond the practical limits of air cooling, reduces fan energy, supports higher coolant temperatures (which expand free-cooling opportunities), and produces higher-quality waste heat that can be recovered for reuse.
Liquid Cooling and ATTOM
Attom Technology developed ByteCool (Direct-to-Chip) and OceanCool (Immersion) specifically to address the thermal wall created by today’s GPU Servers and AI Accelerators. These platforms integrate cleanly into AgileCore modular designs, allowing operators to deploy dense compute while maintaining stable temperatures, improving overall facility efficiency, and opening practical pathways for Heat Recovery.
