Nvidia has put money into Hydra Host, an AI infrastructure startup that operates bare metal GPU clusters across roughly 40 to 50 locations worldwide. The investment comes as Hydra Host positions itself as a key intermediary between GPU supply and the surging demand for AI compute outside traditional hyperscale cloud environments.

The funding picture

Hydra Host closed a Series A round that brought in approximately $70 million, part of a larger $142 million offering that wrapped up on March 10, 2026. The company’s total funding across multiple rounds sits somewhere between $89 million and $148 million, depending on how you count the various stages.

Prior to the Series A, Hydra Host completed a Seed+ round in February 2025 led by Flume Ventures. That earlier raise set the stage for the more substantial capital injection that followed.

The fresh capital is being directed toward expanding GPU infrastructure, with Nvidia’s Blackwell B300 servers playing a central role.

CEO Aaron Ginn has been vocal about the company’s mission to make high-performance AI compute more accessible. The vision is straightforward: take existing data center capacity and convert it into what the company calls “scalable AI production units.”

Why Nvidia cares about this

Nvidia’s involvement goes beyond just writing a check. Hydra Host has been named an official Nvidia Cloud Partner, or NCP. The company is also actively contributing to Nvidia’s DGX Cloud Lepton initiative, which suggests a deeper technical integration than a typical financial investment would imply.

Hydra Host also recently signed a partnership with Duos Technologies, signaling that the company’s client base is diversifying beyond the usual AI startup crowd.