NVIDIA DSX Ready Qualifies Power and Cooling Products for AI Factories
The new program starts with battery storage and cooling distribution units, giving infrastructure buyers a product-level signal while leaving site engineering and stability checks in their hands.
Edited by Tyronne Panaino
NVIDIA launched DSX Ready on September 21 as a qualification program for power and cooling products used in AI factories. The official announcement says the first two categories are battery energy storage systems and cooling distribution units, connecting the company's DSX reference designs to specific partner offerings.
The change matters to data-centre builders, power suppliers and cooling vendors because AI infrastructure design is increasingly constrained by more than compute. A rack plan has to work with available electricity, cooling capacity, water, facilities and grid conditions. DSX Ready gives procurement and engineering teams a product-level compatibility signal, while NVIDIA explicitly leaves site-level engineering and stability decisions outside the qualification.
From a reference design to named products
NVIDIA describes DSX as an AI-factory platform that brings compute, networking, power, cooling, facilities and software into one system design. DSX Ready adds a qualification layer for products that are meant to fit applicable parts of that design.
At launch, the battery-storage category includes solutions from Hitachi Energy, LG Energy Solution and Tesla. The cooling-distribution category includes products from LG Electronics, LiquidStack and Vertiv. NVIDIA says more infrastructure and software categories will be added over time, but the announcement does not provide a schedule or name the next category.
That is a narrower change than a new AI-factory architecture. DSX already supplied the reference-design context; DSX Ready creates a route for particular supplier products to be reviewed against category requirements. The practical delta is discoverability and procurement evidence rather than a new accelerator, model or data-centre deployment.
The qualification paths are category-specific
Battery energy storage and cooling equipment do not follow the same review path. NVIDIA says battery-system providers run required tests and submit supporting data for review and approval within a defined qualification boundary. Cooling-distribution providers use a self-qualification suite to determine whether a product meets applicable functional requirements.
The distinction matters because the badge does not represent one universal test across every component. Buyers need to understand the category, the exact offering and the boundary of the evidence submitted. A product that qualifies for one function is not automatically validated for every facility configuration or operating condition.
For builders, the intended benefit is a shorter path from the reference design to a shortlist of products. For suppliers, the program defines a way to demonstrate that an offering meets the requirements NVIDIA applies to its category. Both benefits are first-party descriptions of the program; the fetched evidence does not include an independent comparison of integration time, failure rates or total project cost.
Qualification is not site certification
NVIDIA states that passing the battery-storage qualification does not replace site-level engineering or imply site-level stability. It makes a similar practical point for cooling: even when a unit meets the relevant functional criteria, the builder still has to determine how it fits the planned facility.
That limitation prevents a common overreading. DSX Ready is not proof that an AI factory using a listed product will meet its energy, thermal, water, uptime or commissioning targets. Those outcomes depend on the complete design, installation, controls, local infrastructure and operating profile.
The program also does not publish the complete test procedures, thresholds, submitted data or comparative results on the announcement page. Procurement teams therefore have a qualification label and category description, but not enough public evidence from this source alone to compare one qualified product with another.
What to watch next
The next useful checkpoints are the addition of new categories, publication of clearer qualification boundaries and field evidence from completed AI-factory projects. Buyers will also need current product lists because qualification attaches to specific offerings rather than to a supplier's entire catalogue.
If DSX Ready becomes widely used, it could make power and cooling choices more legible within NVIDIA-centred AI-factory projects. Whether it reduces integration risk or deployment time remains an outcome to measure, not a result established by the launch announcement.
Status
Confirmed. NVIDIA published the program, initial categories, supplier lists and qualification boundaries. Internal confidence is medium because the evidence is one official company source and no independent deployment results were fetched.
Sources
Update note: Last reviewed 2026-09-28. We will revise this post when NVIDIA adds qualification categories or publishes field evidence about integration and deployment outcomes.
Sources
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.