NVIDIA launches DSX Ready program qualifying power and cooling hardware

By Splitfeed EditorialLast updated Sep 26, 2026, 20:18 UTC

In late September 2026, NVIDIA launched the DSX Ready qualification program, a formal certification framework aimed at power and cooling hardware used in AI data centers, which the company refers to as AI factories. The announcement came across several days, with Crusoe's related Exemplar Cloud validation on the Blackwell Ultra platform confirmed on September 21, followed by the broader DSX Ready program announcements on September 22. In the initial qualification round, at least three battery energy storage systems were selected. Hitachi Energy's BESS and LG's 2.5-megawatt battery system were both named as qualifiers, making them among the first hardware products to carry NVIDIA's new infrastructure endorsement.

The program matters because it represents NVIDIA moving beyond chips and software to exercise influence over the physical infrastructure layer of AI data centers. By setting qualification standards for power storage and thermal management equipment, NVIDIA is positioning itself as an arbiter of what counts as suitable infrastructure for high-density AI workloads built around its hardware. This raises real questions about market dynamics: suppliers that earn DSX Ready status gain a meaningful signal of compatibility and implicit endorsement, while those outside the program may face pressure from data center operators who treat NVIDIA certification as a procurement shortcut. The practical stakes are significant given the enormous power demands of systems built on architectures like Blackwell Ultra.

None of the three camps have published reactions to this story yet. The Pro-AI camp would typically frame a program like this as a necessary and welcome step toward standardizing infrastructure for the next generation of AI compute, reducing deployment friction and accelerating buildout. The Anti-AI camp would likely raise concerns about NVIDIA consolidating control over yet another layer of the AI supply chain, arguing that qualification programs can function as soft gatekeeping that concentrates power in one company's hands. The Middle Ground camp would probably acknowledge the practical utility of hardware compatibility standards while calling for scrutiny of whether the certification process is transparent and accessible to a broad range of suppliers.

The clearest thing to watch going forward is how quickly the qualified vendor list expands and whether NVIDIA publishes the technical criteria behind DSX Ready status. If the standards are openly documented and the program admits a wide range of qualifiers, that would support the infrastructure-standardization framing. If access proves narrow or opaque, concerns about market leverage are likely to sharpen.