Revolutionizing AI Density with
Two-Phase Liquid Cooling

Blackstar Diversified Enterprises partners with Accelsius to deliver NeuCool®, a patented two-phase, direct-to-chip liquid cooling platform that redefines the thermal limits of the modern data center. In the mission-critical sector, the explosion of AI and high-performance computing (HPC) has pushed traditional air cooling to its breaking point. Accelsius provides the industry’s most advanced thermal management solution, utilizing a safe, non-conductive dielectric refrigerant that transitions from liquid to vapor to remove heat with unparalleled efficiency. By combining BDE’s integration expertise with Accelsius’ breakthrough technology, we empower operators to deploy the highest-density GPU clusters while significantly reducing energy costs and eliminating water consumption in the white space.

Our partnership brings "Future-Ready" cooling to the Gulf South and Mid-Atlantic, enabling rack densities exceeding 250kW. Unlike water-based systems, the NeuCool platform offers a zero-risk environment for your most expensive IT assets; in the unlikely event of a leak, the dielectric fluid simply evaporates without damaging sensitive electronics. Together, BDE and Accelsius provide a seamless transition from legacy air-cooled facilities to liquid-cooled AI factories, ensuring that your mission-critical infrastructure remains resilient, sustainable, and capable of supporting next-generation silicon.

Scalable Precision for Gigawatt-Class AI Factories

The BDE-Accelsius collaboration focuses on a suite of modular products designed for rapid scale and ease of maintenance. The NeuCool MR250 serves as our flagship in-row Coolant Distribution Unit (CDU), delivering 250kW of cooling capacity in a compact, redundant architecture that integrates easily into existing facility water loops. For specialized environments, the Multi-GPU vaporators are engineered to cool the most advanced processors (supporting 4500W+ per socket), ensuring that even the most intensive training and inference workloads remain thermally stable.

To ensure absolute confidence in deployment, we offer the Thermal Simulation Rack (TSR) and Load Simulation Sleds (LSS). This "thermal bank" allows our clients to test, validate, and commission their liquid cooling infrastructure in a controlled environment before live GPUs are ever installed. Backed by Accelsius’ North American supply chain and BDE’s comprehensive professional services—including system architecture and on-site deployment—we provide a turnkey cooling solution that maximizes uptime and drives Power Usage Effectiveness (PUE) toward the theoretical limit of 1.0.

MR250 In-Row CDU

Two-phase liquid cooling CDU delivering up to 250kW of rack cooling capacity for high-density AI compute environments.

Key points:
• Row-based CDU architecture
• Supports 1-2 racks up to 250kW
• Two-phase cooling for extreme density

View System

Multi-GPU Cooling Platform

Two-phase cold plate system engineered to support advanced GPU and CPU clusters used in AI and HPC environments.

Key points:
• 4-way and 8-way GPU configurations
• Serial plumbing for simplified integration
• High thermal uniformity across GPUs

View Configuration

TSR Thermal Simulation Rack

Rack-level thermal simulation platform designed to validate liquid cooling systems before production deployment.

Key points:
• Simulates high-density AI workloads
• Integrated instrumentation and monitoring
• Supports up to 250kW thermal load testing

View System