Vultr's AI expansion drives new cloud buildout

Vultr has announced a significant step in its long-running collaboration with AMD as it prepares to roll out an AI supercluster powered by AMD Instinct MI355X GPUs. The company is deploying the GPUs at its new Cloud data centre in Springfield, Ohio. This 50 MW expansion is being positioned as a performance-per-dollar play for training and inference workloads.

Springfield marks a major hardware scale-up

The announcement centres on Vultr’s plan to add 24,000 AMD Instinct MI355X GPUs at the new campus. The company was among the early adopters of AMD Instinct MI325X and MI355X GPUs, and this next phase broadens the infrastructure footprint.

Vultr also signalled future plans to integrate the upcoming AMD Instinct MI450 Series GPUs and the Helios rack-scale infrastructure. The roadmap, as described, builds on a shared direction with AMD to provide high-performance, energy-efficient compute for global customers.

Comments

J.J. Kardwell, CEO, Vultr, said the escalating demand for AI infrastructure continues to shape how the company expands capacity. “As demand for AI infrastructure continues to accelerate, Vultr is committed to delivering hyperscale capacity with speed and global reach,” Kardwell said. He added that investing in racked GPU capacity is intended to help enterprises develop and ship new AI applications faster.

Sharing AMD’s perspective on the collaboration, Andrew Dieckmann, corporate vice president and general manager, Data Centre GPU Business Unit, AMD, said the partnership demonstrates how joint development can lead to large-scale AI infrastructure deployment. “Together, we’re delivering large-scale AI compute that meets the needs of the most demanding AI workloads,” Dieckmann said.

Broader stack integration and regional support

The collaboration between the two companies now extends beyond GPUs. Vultr has also integrated the AMD EPYC 4005 Series processors and the recently introduced Vultr VX1 Cloud Compute into its full-stack infrastructure.

The Springfield data centre marks Vultr’s first Cloud location in Ohio. The project is supported by the Ohio Governor's Office, Department of Development, JobsOhio, Dayton Development Coalition, Greater Springfield Partnership and the City of Springfield. The move adds to Ohio’s push to position itself as a destination for AI and digital infrastructure investments.

A new cluster, a wider strategy

For both companies, the announcement reinforces a shared aim: scaling compute fast enough to keep up with heavier AI workloads. The details point to hardware-led expansion, an emphasis on energy efficiency and a deeper alignment across the full stack. With Springfield becoming part of Vultr’s global network, the deployment sets the stage for the next round of GPU-driven Cloud growth.

Read More:

CrowdStrike on empowering India’s channel partners for cybersecurity’s future