Nvidia launches DSX Ready to certify power and cooling solutions, reflecting a shift from processor speed to limits such as storage and cooling efficiency
Listen to this article
Read by Anchor
Nvidia has launched a new engineering qualification program called DSX Ready, targeting the certification of partner-developed power and cooling solutions to meet the reference requirements of the company’s AI factory system. This move responds to a tangible shift in advanced computing infrastructure management, where expansion is no longer governed solely by processor performance and speed, but electrical grid constraints, storage capacity, water availability, and cooling efficiency now constitute the actual limiting factor for what developers can build and operate in the field.
The new program begins by testing two core categories of industrial hardware: battery energy storage systems (BESS) and liquid cooling distribution units (CDUs). At launch, the certified partners for the energy storage category included Hitachi Energy, LG Energy Solution, and Tesla, while the approved solutions for the cooling distribution category comprised LG Electronics, Liquid Stack, and Vertiv.This approach relies on standardizing cooling and power specifications within a single reference design platform., with announced plans to gradually expand the qualification scope to include additional categories across various infrastructure and software layers.
Nvidia has defined separate audit paths for each technology category. In the energy storage battery sector, manufacturers must conduct specific qualification tests and submit technical data for review and approval by the company's engineers within clearly defined test boundaries. Nvidia emphasized in its announcement that passing these tests does not eliminate the need for specialized site engineering nor guarantee overall site stability. For cooling distribution units, suppliers undergo a self-qualification package to assess how their products meet the functional requirements defined by the platform, allowing teams to evaluate how qualified equipment aligns with facility requirements before making final purchasing and installation decisions.
The announcement carries direct operational relevance for enterprises and data-center operators in the Gulf region and Egypt, where the pace of building super-computing clusters in climates that demand peak cooling efficiency and precise calculations of power and water resources is accelerating. For procurement officers and engineering teams in the region, the program provides a standard that helps reduce technical integration risks when constructing AI clusters, making the use of equipment approved by the chip developer a means to avoid conflicts between server racks and liquid-cooling systems.However, Nvidia's explicit warning that qualification does not replace site engineering places the audit responsibility on the local engineer.This requires teams to verify themselves the readiness of electrical networks and the compatibility of cooling units with the local climate and facility specifications, rather than assuming the certification provides a ready-to-use solution for immediate operation.
The DSX platform embodies an approach that views the AI factory as an integrated system combining processing, networking, power, cooling, and software. This view gains importance from the engineering reality that improving the efficiency of a single subsystem often shifts the bottleneck elsewhere, making alignment of industrial supply chains with reference design standards a necessary step to translate chip capacity into actual, stable computing productivity.