Vultr, described as the world’s largest privately-held cloud infrastructure company, announced plans to launch an Artificial Intelligence supercluster in a new Springfield, Ohio cloud data center. The deployment centers on AMD Instinct MI355X GPUs and is part of a 50 MW expansion the company says will deliver improved performance per dollar for Artificial Intelligence training and inference workloads.
The company said it will add an additional 24,000 AMD Instinct MI355X GPUs to its new Springfield campus. Vultr was among the first cloud providers to deploy AMD Instinct MI325X and AMD Instinct MI355X GPUs, and this latest deployment is presented as a significant scaling of that existing collaboration with AMD (NASDAQ: AMD). Vultr framed the expansion as a way to provide high-performance, energy-efficient Artificial Intelligence infrastructure for customers worldwide.
Looking ahead, Vultr said it will continue to extend its adoption of AMD Instinct GPUs by integrating the next generation AMD Instinct MI450 series GPUs and the ‘Helios’ rack-scale infrastructure. The announcement positions the new Springfield site and its 24,000 GPU addition as part of a longer-term commitment to the AMD Artificial Intelligence platform roadmap and to expanding Vultr’s capability to serve model training and inference demand.
