Vultr Launches 20-MW AI Supercluster in Chennai

Cloud infrastructure provider Vultr has launched a 20-megawatt artificial intelligence supercluster in Chennai, powered by AMD Instinct MI355X graphics processing units. This deployment is Vultr's largest in India and ranks among its two largest AMD AI infrastructure setups globally, alongside a facility in Springfield, Ohio. The initial configuration is expected to house around 10,000 GPUs.

Operations and Expansion

The facility is designed to support large language model training, fine-tuning, and inference for AI labs, startups, enterprises, and research institutions. Vultr plans to expand the facility's capacity in 2027 using AMD's MI455X GPUs and Helios architecture. The company is targeting 600 megawatts of capacity in India by 2028 and a gigawatt by 2029. Global enterprises have already pre-reserved a large portion of the facility's capacity.

Workforce and Power

The Tamil Nadu government has enabled 150 megawatts of connected power at the location, with 20 megawatts active for the initial deployment. Vultr has hired over 100 people in India in the past three months and plans thousands of additional hires in network engineering and operations by March, with the Chennai workforce potentially reaching 5,000.

Context

Chennai has rapidly emerged as a major data centre hub in India, driven by robust state government support and reliable power infrastructure. Large cloud providers are heavily investing in Indian data capacity to meet surging local and global demand for artificial intelligence computing power.