Cloud provider Vultr has made AMD Instinct MI355X GPUs available via its platform.

The GPUs are optimized for AI inference workloads, and are ideal for customers looking to support AI-native workloads.

The AMD Instinct MI335X is built on the 4th Gen AMD CDNA architecture. The chips have 288GB of HBM3E memory, 8TBps bandwidth, and support for advanced datatypes like FP6 and FP4 — accelerating AI training, inference, and scientific simulation.

“With this launch, Vultr is expanding access to next-generation GPUs and setting a new standard for AI infrastructure price-to-performance,” said J.J. Kardwell, CEO of Vultr. “By combining Vultr’s cloud platform with the AMD Instinct MI355X, we are delivering the performance, scalability, and efficiency that enterprises and AI innovators need, and we are making it available globally.”

AMD officially launched the MI355X GPUs on June 12 during its Advancing AI conference in San Jose, California.

“AMD is the trusted AI solutions provider of choice to enable customers to tackle the most ambitious AI initiatives, from building large-scale AI cloud deployments to accelerating AI-powered scientific discovery," said Negin Oliver, corporate vice president of business development, Data Center GPU Business, AMD. "The AMD Instinct MI350 Series GPUs paired with AMD ROCm software provide the performance, flexibility, and security features needed to deliver tailored AI solutions that meet the diverse demands of the modern AI landscape.”

Vultr also offers the AMD Epyc 7003, 4005, and 9004 series processors, and the MI300X and MI325X GPUs.

Earlier this year, Vultr secured $329 million in credit financing to help further expand its AI infrastructure and cloud computing platform.

Oracle is also deploying the MI355X GPUs. Earlier this year, the company announced plans to develop a zettascale AI cluster of 131,072 of the GPUs.