Description
Optimized Energy Efficiency
CubixCore AI-Ready Modular Data Centers implement a comprehensive spectrum of liquid cooling technologies encompassing direct-to-chip, immersion cooling, and rear door heat exchangers to provide exceptional thermal dissipation for high-density AI processors. By integrating cutting-edge liquid cooling with Free Cooling Chillers, the architecture minimizes reliance on mechanical chillers in low-temperature environments, achieving an ultra-low PUE approaching 1.1. Furthermore, the deployment of natural refrigerants, such as R1234ze and R32, substantially reduces the Global Warming Potential (GWP), mitigating environmental impact.
Maximized IT Density and Spatial Efficiency
Rack-based liquid cooling infrastructures deliver up to 100kW of cooling capacity within standard rack footprints, efficiently satisfying the rigorous computational requirements of decentralized AI workloads. This comprehensive suite of liquid cooling solutions is essential for agile facility environments and is seamlessly retrofittable into existing IT racks, optimizing physical space to accommodate aggressive AI infrastructure scaling.
Accelerated Deployment Architecture
Prefabricated within the factory utilizing a skid-based modular framework, the CubixCore AI-Ready system ensures stringent quality control while drastically reducing onsite installation and commissioning timelines. This streamlined methodology accelerates overall deployment by 50%, significantly enhancing market responsiveness. Ranging from all-in-one micro/edge facilities to partial or shell deployments via flexible CubixCore AI container variants, these highly reliable solutions are tailored to meet the dynamic requirements of modern AI learning environments. The robust modular design ensures uninterrupted operational reliability, facilitating continuous AI model evolution and expedited project delivery.
Rack-Based Liquid Cooling
Engineered for AI edge pilots and distributed deployments, this configuration provides a highly compact and thermodynamically efficient solution. It facilitates high-performance computing within standard rack dimensions by delivering precise, localized thermal management. This architecture ensures a flexible and scalable AI infrastructure, which is highly critical for accelerating innovation in decentralized AI applications.
Prefabricated Liquid Cooling
AI training models require immense computational power, subsequently generating extreme thermal loads. Our prefabricated liquid cooling solutions, integrating Direct-to-Chip and RDHx technologies, are specifically engineered to neutralize these thermal challenges. The system's rapid deployment capabilities and inherent scalability empower organizations to aggressively expedite their AI strategic initiatives.
Prefabricated Immersion Cooling
Specifically architected to manage the extreme power densities inherent to large-scale AI training clusters. This solution reliably supports massive computational capacities, ensuring stable and highly efficient high-volume data processing.
Prefabricated Containerized DC
Featuring a highly modular design, this containerized infrastructure enables the rapid provisioning of AI capabilities across diverse geographic locations. It integrates sophisticated, energy-optimized cooling solutions, including closed-loop chilled water systems. Ideal for executing AI deployments in outdoor, edge, and remote environments.
Full Liquid Cooling Solutions for AI
The RDHx system actively absorbs thermal exhaust directly from the servers by replacing the standard rear cabinet door with a precision cooling unit, subsequently returning neutralized, cool air to the facility environment. The heated thermal liquid is then routed to an external heat exchanger, which rejects the accumulated heat from the system. Operating in conjunction with direct-to-chip cooling, the coolant is efficiently circulated via chillers to the backend IT infrastructure.
Precision-engineered cold plates mounted directly on the processing chips absorb thermal output via a liquid coolant, which is then circulated to a dedicated heat exchanger. The thermal load is subsequently transferred to a secondary circuit or an external outdoor cooling tower. The chilled liquid is then continuously looped back into the primary system, ensuring optimal server operating conditions.
IT hardware is entirely submerged within a specialized, non-conductive dielectric fluid, which directly absorbs heat and naturally circulates to an in-row cooling unit or integrated heat exchanger. The primary cooling system then re-circulates the thermally neutralized fluid back into the immersion tank for continuous, highly efficient thermal management.
