Enterprises building AI-native applications are discovering that raw GPU capacity is only half the equation. As Kubernetes users move from cloud-native to AI-native stacks, the question is shifting from who has the most accelerators to who can run distributed, secure workloads at global scale.

The first wave of so-called neoclouds stood up massive GPU clusters for a handful of the largest buyers, but enterprise developers need far more than data center capacity. The Constant Company LLC, which does business as Vultr, has bet on AMD silicon and open software stacks to win AI infrastructure deals. It is positioning itself as a full-stack cloud built to close that gap, according to Kevin Cochrane (pictured), chief marketing officer of Vultr.

“In the early market … it was all about just standing up large GPU clusters wherever you could get land, power and shell, and then just selling it on through to a hyperscaler,” Cochrane said. “And somehow in that mix, people forgot that you still need cloud, right? You still gotta run this thing.”

Cochrane spoke with theCUBE Research’s John Furrier and Rebecca Knight at KubeCon + CloudNativeCon NA, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed why enterprise AI demands distributed cloud infrastructure, how global networking shapes GPU adoption and Vultr’s new programs for open-source builders. (* Disclosure below.)

Why AI-native applications need a distributed cloud

Every computing supercycle reshapes architecture from the data center up to the application layer, according to Cochrane. In AI, the first iteration was tuned for a narrow group of the biggest GPU buyers, he explained. Enterprises have different requirements, and serving them means returning to computing’s longer trajectory.

“We took a little bit of a step backwards because in that early iteration of the new architecture, when we over-indexed on a limited number of buyers, we went back more to centralized systems,” Cochrane said. “And we have to remember that the long arc of compute is towards decentralized distributed systems driven by open standards and driven by open source.”

Distributed AI also changes how inference gets served. Because many users ask the same question, Vultr aims to maximize token reuse through a global namespace that shares KV cache across nodes, work it is pursuing with partners such as Vast Data Inc., Cochrane noted. Connectivity between data centers matters just as much, since enterprises want private networks isolated from every other tenant on a public cloud.

“In the neocloud moment, I think some of the complexity of what’s required to operate workloads at global scale for enterprises has been forgotten or just assumed it wouldn’t matter,” Cochrane said. “But … it’s the biggest barrier to getting enterprises to actually leverage GPUs from a new cloud because the networking connectivity, it just doesn’t work.”

Vultr is also putting resources behind the open-source community building the AI stack. The company has joined the Cloud Native Computing Foundation as a platinum member, and at KubeCon it highlighted its Open Source Credits Program, which provisions infrastructure for the maintainers behind new CNCF projects.

“All of those projects need compute, right? The maintainers aren’t going to suddenly open up their own personal checkbooks,” Cochrane said. “How can you build the next great CNCF project if you don’t have compute?”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of KubeCon + CloudNativeCon NA:

(* Disclosure: Vultr sponsored this segment of theCUBE. Neither Vultr nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE