SUSE, NVIDIA & Vultr launch AI Factory for production

Vultr and SUSE have launched SUSE AI Factory with NVIDIA on Vultr infrastructure, targeting organisations moving artificial intelligence workloads into production.

The offering combines SUSE AI Factory, NVIDIA AI Enterprise software and Vultr's cloud infrastructure in a single validated stack. It is designed to reduce the integration work involved in bringing together AI software, graphics processing units and cloud infrastructure.

At the centre of the platform is a Kubernetes-based architecture designed to run across cloud, on-premises, edge and sovereign environments. That reflects rising demand from businesses seeking greater control over where AI workloads sit, particularly when data residency, regulatory requirements and internal governance rules restrict where models and applications can be deployed.

Vultr, a privately held cloud infrastructure provider, supplies the underlying compute layer, including cloud GPUs, bare metal, Kubernetes, compute, networking and storage. SUSE provides the software layer through its AI Factory platform, while NVIDIA contributes AI software and infrastructure components including NVIDIA NIM, NVIDIA NeMo and NVIDIA Run:ai.

The package includes pre-validated blueprints such as retrieval-augmented generation, GPU orchestration tools and zero-trust security measures. These are intended to help customers avoid building AI environments from scratch as they move beyond pilot projects.

Production push

The announcement comes as more companies try to turn early AI trials into operational systems that can run at scale. Many businesses have experimented with models and prototypes over the past two years, but production deployment has proved harder because it requires integrating software, specialist chips, security controls and infrastructure management.

Kevin Cochrane, chief marketing officer at Vultr, said the new platform is intended to address that bottleneck.

"Enterprises are done experimenting. The real work now is getting AI into production fast, without spending 18 months integrating software, GPUs and infrastructure before a single workload goes live," said Kevin Cochrane, chief marketing officer at Vultr. "SUSE AI Factory with NVIDIA on Vultr infrastructure changes that equation entirely. We have built a validated, full-stack platform that brings together enterprise AI software, NVIDIA acceleration and globally available GPU infrastructure. Organisations can now focus on building AI that delivers results, instead of assembling the stack underneath it. That is how enterprises move from pilot to production with confidence."

SUSE presented the tie-up as an extension of its earlier work on private and governed AI environments. Customers want systems that can be deployed in different locations without being tied to a single public cloud provider, it said.

"When we first introduced SUSE AI Factory with NVIDIA in April, our goal was to deliver a fully governed, open Private Enterprise AI framework, enabling our customers to deploy their intelligence wherever their business demands," said Rhys Oxenham, vice president and general manager of AI at SUSE. "Our partnership with Vultr takes it a step further by providing a turnkey, validated infrastructure stack. Enterprises no longer have to wait or spend months wrestling with integration - they can deploy a GA-ready AI platform on world-class infrastructure today."

Infrastructure focus

The tie-up also underlines NVIDIA's role in linking hardware and software layers in enterprise AI deployments. While demand for accelerated compute remains strong, suppliers are increasingly focusing on packaged systems that combine chips, orchestration and model services with the infrastructure needed to run them.

"Production AI requires more than accelerated compute. It requires enterprise AI software and global infrastructure working together," said Dion Harris, senior director, HPC and AI Infrastructure Solutions GTM at NVIDIA. "By bringing SUSE AI Factory with NVIDIA together with Vultr's global infrastructure, enterprises gain a secure, scalable foundation for moving from experimentation to production."

Industry analysts increasingly argue that infrastructure design, rather than model experimentation alone, will determine whether companies can deploy AI widely across the business. That includes decisions about governance, compliance, inference pipelines and the location of computing resources.

Steven Dickens, chief executive officer and principal analyst at HyperFRAME Research, said the collaboration reflected that shift.

"Enterprise AI success in 2026 is being decided at the infrastructure layer. Organisations need a governed, open substrate that spans from GPU compute all the way up to the inference pipeline, with security and compliance built in from day one. The collaboration between Vultr, SUSE and NVIDIA delivers exactly that. A converged, validated stack that takes enterprises from pilot to production without the integration tax that has held so many back," said Steven Dickens, chief executive officer and principal analyst at HyperFRAME Research.

Vultr said it serves hundreds of thousands of active customers across 185 countries. The company disclosed an equity financing in late 2024 that valued the business at USD $3.5 billion, underscoring investor interest in infrastructure providers seeking to capture AI-related spending outside the largest hyperscale cloud groups.

For SUSE, the launch adds to a broader push to position open-source software and multi-environment deployment as an alternative to more closed, cloud-based AI platforms. The emphasis on avoiding lock-in is likely to resonate with customers seeking flexibility over how and where they run AI applications.

Existing SUSE customers will be able to work with account teams from SUSE and Vultr to assess the platform and scope proof-of-concept projects tied to their own AI workloads.