Direct-to-Chip Cooling Infrastructure

Infractructure ready for Next-Gen NVIDIA GB300 class GPU’s, its predecessors, and accelerator platforms

High-Density AI

Up to 3× higher rack power density

Low Capex

Up to 48% reduction in cooling ‑ related capital costs

Energy Efficiency

Reduce cooling energy consumption by as much as 36%

More White Space

Free up to 62% more usable data‑center floor area

Direct-to-Chip Cooling20kW to 130kW per Rack density

Desert Dragon delivers Direct‑to‑Chip (DTC) cooling: coolant circulated directly across cold plates to extract heat at the silicon level. DTC enables rack power densities air cooling cannot support, optimized for AI, HPC, and hyperscale cloud workloads in KSA.

On Next-Gen modular data center platform has been engineered to provide a flexible, scalable, and resilient deployment path designed to accommodate high-density compute infrastructure supporting even the latest release – NVIDIA GB300 class GPU’s, its predecssors, and accelerator platforms.

Modular Design: DD’s architecture is purpose-built for modular expansion and phased field deployment, enabling staged capacity growth with minimal disruption to existing infrastructure.

Cooling Flexibility: DD’s 4th-generation design natively supports direct-to-chip liquid cooling architectures compatible with leading NVIDIA GPU’s, and other high-density AI accelerator platforms, providing an engineered thermal envelope capable of managing the demands of high-density compute workloads.

Power Infrastructure: DD’s busway infrastructure is designed to meet Tier III topology requirements, delivering concurrently maintainable power distribution with redundant feed paths to support the demands of high-density compute deployments.

Robustness and Operations: Redundant power delivery, modular PDUs, and integrated monitoring are designed to meet Uptime Institute Tier III availability targets (99.982% concurrently maintainable), ensuring continuous operational readiness, simplified maintenance access, and sustained serviceability under live load conditions.

Key DTC advantages:

    • Higher sustained performance from lower silicon temperatures
    • Predictable throughput for training and inference
    • Extended hardware life and reduced failure rates
    • Scalable density to accommodate next‑gen accelerators
    • Hyperscale resilience with enterprise reliability design
    • Reduced total cost of ownership through longer component life and fewer hardware replacements

Proprietary AI Solutions

All Desert Dragon facilities run FinBlade AI, continuously analyzing thermal behavior, predicting demand, and automatically optimizing cooling. The result is a self-optimizing infrastructure built for mission-critical AI, cloud, and enterprise workloads.

Infrastructure Designed for High-Density AI Deployments

Our DTC cooling ecosystem is purpose-built for modern AI and HPC hardware, enabling rapid deployment of high-density server racks without additional engineering. Integrated direct-to-chip liquid cooling allows customers to deploy high-performance hardware without custom cooling integration.

Integrated cooling infrastructure that supports liquid cooled hardware out of the box ​

High density power & cooling delivery for GPU clusters and performance compute ​

Scalable AI infrastructure, from single high power racks to massive AI clusters ​

Flexible deployment options supporting diverse AI/HPC platforms​

Enterprise grade reliability through redundant, monitored cooling systems​

Stable performance for sustained AI and high-performance workloads​

Operational Excellence for DTC Infrastructure

Desert Dragon’s DTC cooling systems are tightly integrated into facility operations and supported by advanced monitoring, maintenance, and management processes. Predictable thermal performance for AI and HPC workloads through continuously monitored cooling systems.

Operational benefits include:

  • Predictable thermal performance for AI and HPC workloads
  • High reliability via redundant pumps and N+1 cooling architecture
  • Real-time monitoring of temperature, pressure, and flow for proactive management
  • Remote oversight through API-enabled management platforms
  • Concurrent maintenance without interrupting compute operations
  • Simplified liquid cooling deployment so customers can focus on compute, not cooling

Our Certifications

Desert Dragon holds industry-leading certifications, ensuring compliance with the highest standards in data center development and operations.