"Vertiv has announced new configurations of its MegaMod HDX solution, a prefabricated power and liquid cooling infrastructure designed for environments with very high power densities. The solution is intended for applications such as artificial intelligence and high-performance computing and is available in North America and the EMEA region. According to Vertiv, the new variants respond to the rapidly growing demand for computing power and associated cooling capacity in data centers."
"The modular design allows operators to add capacity as workloads grow, while keeping implementation time to a minimum. The MegaMod HDX combines liquid cooling directly on the chip with traditional air cooling to handle the heat generated by modern GPU-based systems. Rack densities of over 100 kilowatts The infrastructure is available in two versions. A compact model accommodates up to thirteen racks with a total power capacity of approximately 1.25 megawatts."
"Vertiv emphasizes that the hybrid cooling architecture is designed to handle extreme thermal loads, for example in pod-based AI environments and GPU clusters. In addition, the power supply is set up in such a way that modules can take over from each other in the event of failures, ensuring continuity. An integrated thermal buffer system ensures that systems remain stable during maintenance or sudden changes in load."
MegaMod HDX is a prefabricated modular power and liquid cooling infrastructure aimed at artificial intelligence and high-performance computing environments with very high power densities. The solution combines direct-to-chip liquid cooling with traditional air cooling to manage heat from modern GPU-based systems. Two configurations are available: a compact model for up to thirteen racks (~1.25 MW) and a larger version supporting up to 144 racks and up to 10 MW. Rack densities of 50 kW to well over 100 kW per rack are achievable. The hybrid cooling, redundant power modules, and integrated thermal buffer support stability and continuity. Systems are fully assembled and factory-tested to shorten deployment and improve planning certainty.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]