India’s AI infrastructure race just found its most serious hardware backbone. Yotta Data Services has announced plans to deploy one of Asia’s largest NVIDIA Blackwell Ultra AI superclusters in India, anchored by over 20,000 next-gen GPUs and backed by a capital commitment exceeding $2 billion.
Expected to go live by August 2026, the deployment will sit at the heart of Yotta’s hyperscale data centre ecosystem and positions India firmly among the few global regions capable of hosting frontier-scale AI compute within its borders.
The significance here goes beyond raw silicon. NVIDIA will also establish one of the largest DGX Cloud clusters in the Asia-Pacific region within Yotta’s infrastructure under a four-year engagement valued at over $1 billion. Together, these moves signal a deeper India–US technology alignment and a structural shift in global AI compute supply chains, where trusted regions like India are emerging as distributed hubs for advanced model training and inference.
For India’s sovereign AI ambitions, this means reduced dependence on offshore compute and the ability to build, train, and deploy globally competitive models from infrastructure physically located within the country.
Also read: Zero tax, $200 billion dreams: India wants to power world’s AI infrastructure
At the heart of the announcement is scale – on almost every axis imaginable. Yotta will deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs, forming one of Asia’s largest AI superclusters with a total infrastructure investment exceeding $2 billion. This system is expected to go live by August 2026, marking one of the fastest large-scale AI compute rollouts globally.
The deployment is centred on Yotta’s 60 MW D2 data centre at its Greater Noida hyperscale campus, which is scalable to 250 MW, with additional capacity from its Navi Mumbai campus, designed to scale to a massive 2 GW. Together, these facilities form the backbone of what could become one of the world’s most significant AI compute corridors.
NVIDIA’s role is equally substantial. The company has signed a four-year engagement worth over $1 billion to establish one of APAC’s largest DGX Cloud clusters inside Yotta’s Blackwell-powered infrastructure. This builds on an existing relationship where NVIDIA DGX Cloud has already been utilising Yotta GPU infrastructure over the past year, now expanding in line with surging regional demand.
Also read: Yotta to Adani: India building sovereign, frontier AI with Global South relevance
On the technical front, the supercluster will integrate 800 Gbps NVIDIA Quantum-X800 InfiniBand networking, advanced liquid cooling, and more than 40 petabytes of high-performance parallel storage. It is engineered to handle trillion-parameter model training and multi-million concurrent inference prompts, underscoring its frontier-scale ambitions.
Crucially for domestic ecosystem development, Yotta is committing over 10,000 NVIDIA B300 GPUs from the supercluster to the IndiaAI Mission, enabling sovereign foundation model development and research.
This is all in addition to what Yotta already operates in India – 10,000+ NVIDIA GPUs in production, with another 8,000 GPUs going live within the next quarter. By FY27, Yotta plans to scale beyond 80,000 NVIDIA GPUs, with a long-term roadmap capable of exceeding one million GPUs over the next three to five years.
“India’s AI ambition requires sustained, high-performance compute at scale,” said Sunil Gupta, Co-Founder, MD & CEO, Yotta Data Services. NVIDIA’s Jensen Huang framed the strategic significance succinctly: “India is emerging as one of the world’s most important AI markets.”
This announcement isn’t a surprise, just long overdue. If data is the new oil, then compute is the refinery. And India, it seems, is building both.
Also read: Sarvam to Yotta: NVIDIA shows India AI ecosystem scale
