NVIDIA on September 21 introduced DSX Ready, a new qualification program that vets partner-made power and cooling hardware against the company’s DSX AI factory reference design, giving builders a defined list of pre-approved products for their infrastructure deployments.
What DSX Ready qualifies and why it exists
NVIDIA is introducing DSX Ready as a qualification program for partner products and solutions that meet applicable NVIDIA DSX AI factory reference design requirements. The program launches with two initial categories: battery energy storage systems (BESS) and cooling distribution units (CDUs).
The NVIDIA DSX AI factory platform unifies AI factory design and operations across compute, networking, power, cooling, facilities and software, helping partners design and operate the factory as one system to produce more useful AI output within available power, cooling, water and grid constraints.
That system view matters because optimizing one part of an AI factory can shift the bottleneck elsewhere, meaning power and cooling suppliers need a clear path from reference designs to qualified offerings. Builders need a clearer way to discover and evaluate those offerings.
The program aims to reduce integration risk for AI factory builders by giving them a clear qualification standard to evaluate offerings, though passing qualification does not replace site-level engineering.
Six founding partners across two hardware categories
The first two categories are battery energy storage systems, where Hitachi Energy, LG Energy Solution and Tesla made the initial list, and cooling distribution units, where LG Electronics, LiquidStack and Vertiv qualified.
For battery energy storage systems, partners run required qualification tests and submit supporting data for NVIDIA review and approval within a defined qualification boundary; passing qualification does not replace site-level engineering or imply site-level stability. For cooling distribution units, the path uses a self-qualification suite to determine whether a specific offering meets applicable NVIDIA functional requirements; teams can then focus on how qualified offerings fit their site, configuration and operating needs. CDUs form part of the liquid-cooling infrastructure used to move heat away from increasingly dense AI systems, and NVIDIA’s reference guidance calls for coordinated designs covering flow rates, redundancy and facility integration.
A DSX Ready qualification indicates that a product meets the relevant NVIDIA criteria, while operators must still determine whether it fits the engineering requirements of a particular site. Additional categories across infrastructure and software will be rolled out over time.
DSX platform context
NVIDIA announced the broader DSX platform at GTC Taipei, describing it as a complete playbook to create AI factories, bringing together open source, modular software libraries, application programming interfaces, reference designs, NVIDIA accelerated computing platforms and partner technologies into a common, codesigned platform for AI factory design, deployment and operations.
DSX MaxLPS reduces per-rack power requirements and enables up to 40 percent more GPU capacity within the same facility envelope, with performance-per-watt improvements of 1.3, 1.5x, contingent on optimized infrastructure design, workload profiling and early site-level engagement for 45°C liquid cooling. DSX OS is open source modular software for AI factory operators to provision, operate and monitor AI infrastructure at scale. The DSX Ready program announced this week extends that platform logic into the vendor-selection process, connecting specific third-party hardware to the reference designs builders are already working from.
Power management and energy efficiency are increasingly important to the economics of AI infrastructure investment, as NVIDIA projects that 100 GW of AI factory capacity will come online globally before the end of the decade, making facility power budgets a primary constraint on how much computing capacity operators can deploy.
That constraint is precisely what makes a standardized vendor qualification list commercially significant. Builders of large-scale facilities cannot afford to evaluate every potential supplier independently, and the DSX Ready catalog shortens that process considerably.
Market implications for infrastructure vendors
The program arrives as power availability and thermal management become larger constraints on AI infrastructure expansion, with NVIDIA developing DSX around those limitations, including new power architectures for high-density systems and technologies intended to make AI factories more responsive to electricity supply conditions.
Suppliers that clear the DSX Ready bar gain visibility to every builder working from NVIDIA’s reference designs, while those outside the program risk being passed over on reference-design bids. That commercial pressure is likely to accelerate partner submissions as NVIDIA rolls out additional product categories.
The qualification program is detailed on the official NVIDIA blog. For a broader look at how AI infrastructure spending is reshaping the semiconductor supply chain, see GCN’s earlier report on how one chipmaker captured 46 cents of every server CPU dollar in early 2026. Builders wrestling with the financial architecture of AI deployments may also find relevant context in GCN’s breakdown of hidden egress costs in cloud infrastructure.
