Digi Power X pitches enterprise AI computing capacity built on Supermicro servers with Nvidia Blackwell chips

Digi Power X Inc is positioning itself as a provider of enterprise artificial intelligence computing capacity by combining its own power infrastructure with GPU servers built by Super Micro Computer Inc, according to a report published by Proactive Investors and syndicated via Yahoo Finance.

The company’s approach centers on using Supermicro systems equipped with Nvidia’s latest Blackwell chips, then selling the resulting compute capacity to enterprise customers, the report said. The emphasis is on delivering usable AI workloads rather than only selling components, which puts the business closer to a “capacity as a service” model than a traditional hardware reseller.

Blackwell is Nvidia’s newer data-center platform designed to accelerate AI training and inference, and it is widely used as a benchmark for next-generation GPU performance. In this arrangement, Supermicro acts as the server OEM, supplying the physical compute nodes, while Digi Power X supplies the surrounding infrastructure, starting with power.

A key theme behind the report is infrastructure readiness. In AI deployments, power availability and speed of ramp matter as much as GPU choice, because high-density computing quickly creates constraints for data centers. By highlighting that it “pairs” owned power infrastructure with GPU-equipped servers, Digi Power X is effectively selling a faster path to bringing AI systems online.

Nvidia, meanwhile, supplies the processors. Its role in this chain is the provision of the GPU architecture used inside the servers, while Supermicro provides server hardware designed to support those GPUs. For Nvidia, being named in third-party system builds underscores the breadth of its data-center ecosystem, particularly as enterprises look to scale deployments through multiple OEM partners.

The Proactive Investors write-up did not provide details on Digi Power X’s customer contracts, contract sizes, minimum commitments, pricing, or timelines. It also did not disclose whether the company operates the compute as a managed service with software layers, or whether customers install their own workloads after capacity is delivered.

As for what is still unclear, the report also did not explain the degree of Digi Power X’s “owned power” footprint, such as location, capacity, redundancy design, or how it manages utilization rates as demand changes. Those operational details typically determine whether an infrastructure provider can reliably deliver AI capacity at scale.

Looking ahead, investors and enterprise buyers are likely to focus on whether Digi Power X can expand capacity smoothly as Blackwell-based systems roll out, and whether the company can convert its infrastructure advantage into repeatable enterprise demand, especially in a market where supply chains and data-center power planning can be major bottlenecks.