Cloud Infrastructure Firm Vultr Expands AI Supercluster in Chennai with AMD GPUs

Vultr, a privately held cloud infrastructure company, has announced the expansion of its AMD-powered infrastructure designed for enterprise and sovereign cloud initiatives. The company is deploying AMD Instinct MI355X GPUs in a new 20-megawatt supercluster located in Chennai, India.

This new facility establishes the largest AMD Instinct deployment in the country and positions Vultr with two of the world's largest AMD AI infrastructure deployments globally, alongside its existing flagship supercluster in Springfield, Ohio.

The Chennai supercluster is targeted at serving a wide range of users, including AI labs, foundation model builders, AI-native startups, traditional enterprises, as well as research and academic institutions. The infrastructure is built to support demanding artificial intelligence workloads, such as large language model and foundation model training, fine-tuning, and high-throughput inference. It is intended to provide organizations with the computational capacity required to develop and deploy AI at scale.

The underlying hardware features AMD Instinct MI355X GPUs equipped with 288GB of HBM3E memory and 8TB/s of memory bandwidth per GPU. This configuration supplies the necessary performance and memory capacity to handle complex AI models.

J.J. Kardwell, CEO of Vultr, noted the operational needs of local deployment. Enterprises deploying AI at scale require infrastructure they can access locally to meet data sovereignty, data residency, and latency requirements. Kardwell stated that the Chennai supercluster gives organizations in India the capacity to train, fine-tune, and run inference on in-country infrastructure.

Additionally, the supercluster will leverage the AMD ROCm open software ecosystem. This allows developers to access widely used AI frameworks, libraries, and models, providing flexibility in how they develop and deploy AI workloads.

Jaya Jagadish, Country Head and SVP Engineering at AMD India, commented on the local ecosystem, stating that India possesses the talent and ideas to build the next generation of AI, and that access to compute at this scale will help translate those ideas into reality.

"The expansion of advanced GPU-powered cloud infrastructure in India marks a notable step for the domestic tech ecosystem. As AI adoption accelerates among enterprises and startups alike, having local access to high-performance compute capacity addresses critical data residency and latency requirements. This development enables Indian AI builders to train and deploy complex models efficiently within the country, supporting the growth of sovereign cloud initiatives and advanced technological research." — Dr. Shishir Gupta, Founder & CEO, StartupLanes