India’s AI infrastructure is about to gain a major new engine. On 21 Oct 2026, Vultr and AMD announced an expanded collaboration to deploy a 20 MW supercluster in Chennai, powered by AMD Instinct MI355X GPUs and built for organizations that need serious compute inside the country.
The project will become India’s largest AMD Instinct deployment, giving AI labs, foundation model builders, AI-native startups, enterprises, and research and academic institutions access to a powerful in-country platform. What happens when more AI teams can train, fine-tune, and run their models without sending workloads beyond India’s infrastructure? Chennai is set to provide part of that answer.
AMD’s MI355X GPUs Drive Chennai’s Expansion
Vultr will deploy AMD Instinct MI355X GPUs in the Chennai supercluster, with each GPU offering 288GB of HBM3E memory and 8TB/s of memory bandwidth. Those specifications give the system the capacity to handle demanding AI workloads across training, fine-tuning, and high-throughput inference.
The supercluster is designed for large language model and foundation model training, along with the later stages that turn trained systems into usable products. Teams will be able to fine-tune models for specific needs and run inference at high throughput, creating a shared foundation for AI development across startups, companies, laboratories, and academic institutions.
The Chennai deployment will also leverage the AMD ROCm open software ecosystem. That connection brings the hardware and software pieces together as Vultr expands infrastructure built around AMD’s AI platform.
Vultr’s relationship with AMD spans the AMD Instinct GPU roadmap, including MI300X, MI325X, MI355X, and planned MI455X GPUs. It also includes AMD Helios rack-scale architecture and AMD EPYC data center CPUs, creating a collaboration that reaches from individual processors to larger AI systems.
A Global Cloud Platform With India at the Center
Chennai is not Vultr’s first major AMD AI infrastructure project. Vultr has two of the world’s largest AMD AI infrastructure deployments globally, including its supercluster in Springfield, Ohio. The Chennai project extends that footprint into India while reflecting a shift toward geographically distributed AI infrastructure.
Vultr provides secure, compliant infrastructure at a predictable cost in 17 countries across six continents, supported by 34 global cloud data center regions. That network gives the Chennai supercluster a place inside a wider infrastructure strategy rather than treating it as an isolated build.
Demand for sovereign cloud infrastructure is rising globally because of security concerns, regulatory requirements, and performance expectations. For organizations developing AI in India, in-country infrastructure can provide a path to keep compute closer to the teams and workloads that depend on it.
J.J. Kardwell, CEO of Vultr, described the goal in direct terms: “Vultr’s Chennai supercluster, powered by AMD MI355X GPUs, gives organizations in India the capacity to train, fine-tune, and run inference on in-country infrastructure.”
That focus on local access matters because the supercluster is aimed at a broad AI community, not one narrow customer group. AI labs, foundation model builders, AI-native startups, enterprises, and research and academic institutions will all sit within the project’s intended audience.
The Next Step Arrives in 2027
Vultr plans additional capacity expansion in 2027, including AMD Instinct MI455X GPUs and AMD Helios rack-scale AI architecture. The planned expansion points to a continuing buildout as Vultr and AMD extend their relationship across new generations of AI hardware and rack-scale systems.
For India’s AI ecosystem, the significance reaches beyond one facility. “India has the talent and ideas to build the next generation of AI,” said Jaya Jagadish, Country Head & SVP Engineering, AMD India. “Access to compute at this scale will help turn more of those ideas into reality.”
Vultr was founded by David Aninowsky and has been self-funded for over a decade. In December 2024, Vultr announced an equity financing at a $3.5 billion valuation, adding a major financial milestone to the company’s expansion story.
Now the Chennai supercluster gives that expansion a clear AI infrastructure focus. A 20 MW site, AMD Instinct MI355X GPUs, ROCm software, and planned MI455X and Helios capacity will place India inside a growing network of distributed compute. As AI builders look for secure, compliant, and predictable infrastructure, Chennai is preparing to become one of the most important AMD-powered centers in that network.
Based on
