Privately-held cloud company Vultr has deployed AMD Instinct MI325X GPUs in its Chicago data center region.
The company claims to be the first cloud provider to make the GPUs available to customers.
AMD revealed its MI325X GPU in October 2024. Built on CDNA 3 architecture, the AMD Instinct MI325X has 256GB of HBM3E capacity and 6TBps of memory bandwidth. According to the company it offers 1.8x more capacity and 1.3x more bandwidth than Nvidia's H200 GPU, and has 1.3x the peak theoretical FP16 and FP8 compute performance of the H200.
Vultr is offering the GPUs via Supermicro AS-8126GS-TNMR 8-U servers.
According to Vultr, the MI325X is compatible with the MI300X and easily integrates with AMD ROCm software.
“The AMD Instinct MI325X sets new AI industry standards, delivering incredible performance and efficiency for inference,” said J.J. Kardwell, CEO of Vultr. “The Instinct MI325X gives our customers priority access to state-of-the-art GPU hardware, empowering them to scale AI deployments in production. Together, we are unlocking new possibilities for AI innovation and shaping the future of cloud GPU infrastructure on a global scale.”
“Designed for exceptional performance and efficiency, the AMD Instinct MI325X accelerators excel in demanding AI tasks such as foundation model training, fine-tuning, and inferencing,” added Negin Oliver, corporate vice president of business development, data center GPU business unit, AMD. “With the addition of the AMD Instinct MI325X, Vultr is empowering businesses with cutting-edge AI technology to harness the necessary computational power and efficiency to drive AI advancement and accelerate growth.”
In December 2024, Vultr completed a $333 million funding round which was led by LuminArx Capital Management and AMD Ventures. December also saw the company launching an AMD GPU supercomputer cluster at its Chicago cloud data center region in collaboration with AMD, Broadcom, and Juniper Networks.
Vultr has been offering AMD's Instinct MI300X GPUs via its AI cloud since September 2024.
