At AI Summit, Yotta CEO Ignites Sovereign AI Dreams With NVIDIA Collaboration

Since its founding in 2019, Yotta Data Services has rapidly emerged as a key player in India’s AI infrastructure landscape. In 2023, it first partnered with NVIDIA to build Shakti Cloud supercomputer with 16,000+ NVIDIA H100 Tensor Core GPUs. The company now boasts holding 75% of India’s AI GPU compute capacity.

Now, it looks to accelerate the GPU capacity with an expanded NVIDIA partnership.

Yotta announced it will deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs to form one of Asia’s largest AI superclusters, backed by an investment exceeding $2 billion, with deployment expected to go live by August 2026. 

The supercluster will be housed at Yotta’s 60 MW D2 data centre in Greater Noida, scalable to 250 MW, with additional capacity from its Navi Mumbai campus.

Speaking to AIM at the sidelines of the India AI Impact Summit, Sunil Gupta, Co-founder, Managing Director and CEO of Yotta Data Services, lays out an ambitious roadmap that could position India as one of Asia’s largest AI compute hubs.

Gupta frames the announcement as part of a broader shift in India’s AI journey, from compute scarcity to sovereign capability. “Yotta is demonstrating our compute capability, which we have built in India,” he says. 

“From not having any GPUs to [having] enough compute available [with] India announcing so many sovereign models, it’s a very satisfying feeling when I say that these models have all been trained on [our] compute infrastructure.”

Yotta currently operates over 10,000 advanced NVIDIA GPUs in production, with another 8,000 GPUs going live next quarter, before the Blackwell Ultra deployment. Gupta notes that many Indian foundation models already in production have been trained on Yotta’s infrastructure. 

Yotta’s NVIDIA Blackwell supercluster uses NVIDIA’s reference architecture, featuring 800 Gbps Quantum-X800 InfiniBand networking, advanced liquid cooling, and over 40 petabytes of high-performance storage. It is designed to train trillion-parameter foundation models and to handle high-throughput inference with millions of simultaneous prompts.

The CEO boasts that Yotta's AI Factories ensure fast market deployment, low token costs, and high returns. AI developers, enterprises, and governments can rely on this infrastructure for reliable support for large-scale training and critical inference tasks.

“These models are getting into production. They are serving various use cases at population scale,” Gupta remarks, adding that the next phase would require a quantum leap in compute.

The company has also signed a four-year engagement with NVIDIA, valued at over $1 billion, to establish one of APAC’s largest DGX Cloud clusters within Yotta’s supercluster. NVIDIA’s DGX Cloud has already been utilising Yotta’s infrastructure, and the expanded partnership deepens that relationship.

Gupta described the engagement as a milestone for India’s cloud ecosystem. “Possibly for the first time, a global company is giving a very large platform contract, a compute cloud contract, to an Indian company,” he adds. “They feel assured that an Indian cloud company will be able to give very high-performance cloud infrastructure for serving the global and APAC markets.”

Beyond infrastructure scale, Yotta is integrating NVIDIA’s AI software stack, including Nemotron open models and NIM microservices, into its Shakti Studio platform. The company has committed over 10,000 GPUs from the new supercluster to the IndiaAI Mission, supporting sovereign model development and public-sector AI platforms.

“Jensen has publicly declared that they will be open-sourcing the Nemotron model and the entire software stack. The entire Nemotron model and software stack already runs on Yotta GPU clusters, and it will continue to run on the new B300 cluster. People will be coming and consuming it,” the entrepreneur adds.

For Gupta, sovereignty is central to the strategy. “Because Yotta has been doing infrastructure using the best technology, but sovereign, which is in our control, people were looking for this combination,” he says, pointing to growing investor interest.

Yotta is currently in a pre-IPO fundraising phase, targeting high-net-worth individuals and family offices. “The response is phenomenal,” Gupta vaunts.

In total, Yotta aims to raise approximately $1 billion, with $500 million expected from family offices and high-net-worth individuals, and an additional $500 million anticipated from its upcoming IPO. 

If it secures $1 billion, the company believes it will be sufficient to cover its current share of GPU procurement cycles. Yotta is optimistic that by then, market response and GPU adoption will be large enough for all GPU investments to begin generating revenue. This revenue is anticipated to sustain the AI infrastructure company’s future initiatives. 

Gupta’s ambitions stretch well beyond the 20,736 Blackwell GPUs. With existing and planned deployments, Yotta expects to reach nearly 40,000 GPUs, and Gupta says that is only an interim milestone. “I’m not going to stop,” he says. “My target for this year is almost 80,000 GPUs. My capacity to hold the GPUs is already there. Between Mumbai and Delhi, the buildings are ready.”

In light of the expanded partnership, Jensen Huang, Co-founder and CEO, NVIDIA, said in a statement, “India is emerging as one of the world’s most important AI markets, driven by extraordinary talent and a bold national vision. Expanding AI Factory capacity in India strengthens NVIDIA’s regional footprint while supporting India’s ambition to build secure, sovereign, and globally competitive AI.”

At scale, the trajectory reflects India’s evolving role in the global AI value chain. With combined capital commitments exceeding $3 billion across infrastructure deployment and contracted DGX Cloud capacity, Yotta’s expansion signals sustained demand for high-performance AI compute in the region.

Darshan Hiranandani, Co-founder and Chairman, Yotta Data Services, said, “AI infrastructure is becoming foundational economic infrastructure…Our capital strategy is focused on building scalable infrastructure that serves both national priorities and international AI demand.”

Gupta emphasises that the goal is to raise industry awareness of Yotta’s readiness for the next wave of adoption, in which India will embrace these advancements at scale. Furthermore, he notes that this initiative could lead to broader global use of India's computing resources.