Yotta Data Services has announced plans to build one of Asia’s largest artificial intelligence (AI) data center computing hubs in India, backed by an investment exceeding $2 billion and powered by cutting-edge chips from Nvidia. The facility will use Nvidia’s latest Blackwell Ultra GPUs and host a massive AI supercluster intended to serve enterprise and hyperscale customers across the Asia-Pacific region.

The project reflects a rapid escalation in AI infrastructure investment globally. This is as cloud majors including Microsoft and Amazon expand data center capacity in India to meet surging generative-AI demand and data localization requirements.

Indian tech firm Yotta Data Services plans to build a $2 billion data center that will use Nvidia chips also comes amid geopolitical shifts in semiconductor supply chains driven by U.S. export controls. The changes are which are pushing technology companies to deepen partnerships in markets like India.

India is also emerging as a major regional hub for hyperscale data centers with investments like the $7 billion data center hub by Amazon in Telengana. Additionally, Navi Mumbai alone accounts for about 44% of national capacity and hosts projects from multiple global operators.

Project scope and technical details for Yotta data center in India

Total investment: More than $2 billion

Hardware: 20,736 liquid-cooled Blackwell Ultra GPUs

Cluster: One of the largest DGX Cloud deployments in Asia-Pacific

Launch target: August 2026

Core site: Greater Noida hyperscale campus near New Delhi

Additional capacity: Mumbai data-center infrastructure

Additionally, the Noida campus alone hosts a 60 MW D2 data center scalable to 250 MW, while the Mumbai campus can scale to 2 GW.

Financing structure

The financing approach for India tech firm Yotta Data Services to build a $2 billion data center that will use Nvidia chips combines direct capital expenditure with long-term vendor partnership commitments.

Total project capital: More than $2 billion

Supply agreement: More than $1 billion four-year deal with Nvidia to deploy DGX Cloud infrastructure and GPUs

Technology-as-service component: Nvidia will run part of the GPU capacity to serve its own global clients

This vendor-financed model also shows how hyperscale cloud procurement strategies allow rapid deployment of AI compute data without full upfront hardware ownership.

Project factsheet for Yotta AI supercluster data center in India

Location: Greater Noida and Mumbai, India

Developer: Yotta Data Services

Parent Company: Hiranandani Group

Technology Partner: Nvidia

Investment: $2 billion

GPU Count: 20,736 Blackwell Ultra GPUs

Cluster Type: DGX Cloud supercluster

Launch Target: August 2026

Primary Facility: 60 MW D2 data center scalable to 250 MW

Secondary Capacity: Navi Mumbai campus scalable to 2 GW

Contract Value: $1 billion Nvidia partnership for 4 years

Purpose: Support high-performance AI compute and enterprise cloud workloads

Project team and stakeholders

Core Stakeholders

• Owner/Developer: Yotta Data Services
• Parent group: Hiranandani Group, a real-estate conglomerate backing Yotta.

Strategic Technology Partner

• Nvidia for GPU supply, DGX Cloud cluster deployment, AI software ecosystem.

Infrastructure Footprint Contributors

• Greater Noida hyperscale campus as the primary deployment site.
• Mumbai campus for supplementary compute capacity.

Platform Layer

• Yotta’s Shakti Studio AI platform, providing enterprise access to Nvidia’s software stack.