DARKNX cooling systems are specified around the actual thermal outputs of the compute hardware being deployed, not around a generalised standard. Every rack density target is designed to, not designed around.
Chiller units are engineered to DARKNX specification and built by qualified OEM partners under our schedule authority, giving tight integration with the broader thermal and power architecture without exposure to open-market cooling lead times.
All DARKNX cooling designs include N+1 minimum redundancy at the cooling plant level, with topology options up to 2N for mission-critical deployments where no single failure can impact availability.
Not every deployment demands two-phase cooling. DARKNX designs, specifies, and delivers single-phase liquid cooling and precision air-cooled environments for projects where density targets, site conditions, or operational requirements call for proven, cost-effective thermal infrastructure.
Single-phase liquid cooling uses chilled or ambient-temperature water flowing through cold plates and rear-door heat exchangers to remove heat from compute hardware. DARKNX integrates single-phase systems across greenfield, brownfield, and modular deployments, including facilities that require a phased migration path from air to full liquid cooling.
Air cooling remains the right solution for many deployments, particularly enterprise colocation, edge computing, and lower-density compute environments. DARKNX designs precision air-cooled data halls with hot/cold aisle containment, CRAC/CRAH units, and computational fluid dynamics (CFD) modelling to maximize efficiency before liquid cooling is required.
For workloads requiring 100kW+ per rack, explore our Two-Phase Direct-to-Chip cooling solutions in partnership with Accelsius.