Cloud infrastructure company Vultr announced Tuesday it will invest more than $1 billion to launch a new artificial intelligence cluster powered by Advanced Micro Devices technology. 
AMD Processors at a new data center in Springfield, Ohio.
The new cluster is a 50-megawatt, 24,000-chip facility that deploys AMD’s Instinct MI355X graphics processing units and is interconnected with an Ethernet-based fabric that increases data transfer speeds and traffic flow.
Vultr aims to benefit from the AI boom by positioning itself as a cost-effective alternative to hyperscalers and delivering compute at lower prices. Vultr CEO JJ Kardwell said the company’s cloud infrastructure services are typically half the price of hyperscalers.
The company expects the new cluster to be operational by early 2026. “We expect clusters to be sold before going live. We also have MI355X GPUs in racks in the Chicago area for customer testing and piloting,” Cardwell said in an interview with Reuters.
In December 2024, Vultr secured a valuation of $3.5 billion in a funding round led by LuminArx Capital Management and AMD Ventures, the investment arm of AMD.
The company said it does not have any customers in the Ohio cluster yet, but is in active discussions with several companies. Some existing customers, including Chai and LiquidMetal AI, are currently leveraging Vultr’s AMD-powered products in other clusters.
Vultr acquires chips from companies like AMD and Nvidia and operates them from leased data center space.
The company has been providing access to central processing units since 2014. Over the past two years, AI infrastructure has become a significant contributor to Vultr’s revenue since integrating GPUs into its services in 2021.
Vultr, based in West Palm Beach, Fla., is financing the cluster through extended financing facilities with a syndicate that includes Bank of America, JPMorgan, Goldman Sachs and Wells Fargo.
