Yotta to spend $2B on NVIDIA chips

Yotta Data Services on Wednesday announced over a $2 billion spend on NVIDIA's latest chips in an artificial intelligence computing hub it is setting up just outside the national capital.

In a statement, the firm said it plans to deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs (or chips) by August. The deployment will "position India among a select group of countries capable of hosting frontier-scale AI infrastructure."

The US firm will use half of the chips itself over four years for its DGX AI cloud services, which are employed by tech firms such as Tata Consultancy Services and Infosys.

"The collaboration reflects a broader shift in global AI compute supply chains, where advanced AI infrastructure is increasingly distributed across trusted regions.

"India's emergence as a major AI infrastructure node reinforces strategic technology collaboration between India and the United States and strengthens shared priorities around secure, high-performance AI ecosystems," Yotta said.

Backed by a real estate group headed by Niranjan Hiranandani, Yotta runs three data centre campuses in Mumbai, Gujarat, and Greater Noida near Delhi.

The new AI supercluster will be deployed at a hyperscale data centre in Greater Noida. It is designed to support trillion-parameter foundation model training and high-throughput inference workloads handling millions of simultaneous prompts.

Yotta will also allocate over 10,000 NVIDIA B300 GPUs from the cluster to the IndiaAI Mission to support sovereign model development, research institutions, startups and public AI platforms.

Darshan Hiranandani, Co-Founder and Chairman, Yotta Data Services, said, "AI infrastructure is becoming foundational economic infrastructure. This NVIDIA Blackwell Ultra supercluster reinforces India's position in the global AI value chain."

Sunil Gupta, Co-Founder, MD and CEO, Yotta Data Services, added, "India's AI ambition requires sustained, high-performance compute at scale."

NVIDIA Co-Founder and CEO Jensen Huang said, "India is emerging as one of the world's most important AI markets... Expanding AI Factory capacity in India strengthens NVIDIA's regional footprint while supporting India's ambition to build secure, sovereign, and globally competitive AI."

Yotta currently operates over 10,000 NVIDIA GPUs, with another 8,000 set to go live next quarter, and plans to scale beyond 80,000 GPUs by FY27 as India expands its AI infrastructure footprint.

Recently, Neysa, an AI cloud provider, announced a capital raise totalling $1.2 billion. The investment will enable Neysa to deploy more than 20,000 specialised graphics processing units (GPUs) across the country. These purpose-built systems are essential for organisations that need to train and run complex AI models.

Meanwhile, globally, Big Tech firms are locked in an AI infrastructure arms race, pouring hundreds of billions into data centres, custom chips, and power in a long-term bet on competitive advantage that brings rising risks of overcapacity, energy constraints, and investor scrutiny.

(With inputs from PTI)