An Indian Company Is Set to Build a $2 Billion AI Hub With Nvidia’s GPUs and Go Public. Here’s What We Know So Far

India’s artificial intelligence infrastructure landscape is undergoing a dramatic transformation. Yotta Data Services, a homegrown data center company, has announced an ambitious $2 billion investment to build one of Asia’s largest AI superclusters, positioning itself as a critical player in the global AI compute race while preparing for a public market debut.[1][2]

The Scale of the Ambition

Yotta’s investment represents a watershed moment for India’s AI ecosystem. The company plans to deploy more than 20,000 Nvidia Blackwell Ultra GPUs across its facilities, with the infrastructure expected to go live by August 2026.[1] This deployment will be housed primarily at Yotta’s Greater Noida hyperscale campus, supported by operations in Navi Mumbai, creating a distributed but integrated AI compute powerhouse.[1]

To put this in perspective, Yotta currently operates around 10,000 GPUs and already commands 60% to 70% of India’s total GPU capacity, making it the dominant player in the country’s AI infrastructure landscape.[2] The new investment will effectively double or triple its existing capacity, fundamentally reshaping the competitive dynamics of AI compute in South Asia.

Why This Matters for India’s AI Future

For years, India’s AI sector has faced a critical bottleneck. According to Sunil Gupta, co-founder and CEO of Yotta, India possessed the talent, data, and startup ecosystem necessary for AI development—but lacked sufficient compute infrastructure.[1] This gap threatened to limit the country’s ability to develop and deploy large-scale AI models domestically.

The $2 billion investment directly addresses this constraint. As Indian-built AI models serving agriculture, healthcare, and education move into production and scale to serve millions or billions of users, the infrastructure demands become staggering.[1] Gupta emphasizes that while much industry attention has focused on training frontier models, the real computational bottleneck lies in inferencing—where trained models generate content, predictions, and decisions for end users.[1] Serving population-scale applications requires “thousands and thousands of GPUs purely for inferencing,” according to Yotta’s leadership.[1]

A Regional and Global Play

Yotta’s ambitions extend far beyond serving India alone. The company is positioning itself to become a regional compute hub for the Asia-Pacific region and the Global South.[1] Yotta is already in discussions with multiple countries to explore whether their AI capacity could be delivered from Indian data centers.[1]

This regional strategy aligns with broader trends in global AI infrastructure development. OpenAI, for instance, has signed a 100 MW data center capacity agreement with Tata Consultancy Services in India, with plans to scale up to 1 GW, demonstrating that major global AI companies see India as a strategic location for compute infrastructure.[2]

The IPO Path Forward

Perhaps most significantly for investors, Yotta is preparing for a public market listing. The company plans to fund additional GPU purchases through a $1.2 billion to $1.5 billion pre-IPO funding round and aims to list on public markets within the next 12 months.[2] This timeline suggests that Yotta’s leadership has substantial confidence in the investment thesis and the trajectory of India’s AI infrastructure market.

The pre-IPO round valuation and the company’s aggressive expansion plans indicate that investors see substantial long-term value in India’s AI compute opportunity. Yotta’s existing market dominance—controlling nearly three-quarters of India’s GPU capacity—provides a strong foundation for this public debut.

Market Demand Driving Growth

The surge in demand for AI compute is not theoretical. Nvidia’s data center revenue reached $62.13 billion, surpassing expectations, and the company anticipates revenue of $78 billion for the upcoming fiscal quarter.[2] This explosive growth in GPU demand is being driven by enterprises worldwide racing to deploy AI applications, and India is increasingly becoming a critical node in this global infrastructure network.

Local AI model development is accelerating this demand. Multiple companies have launched Nvidia GPU-based AI models at recent AI summits in India, signaling that the country’s AI ecosystem is maturing rapidly and driving genuine infrastructure requirements.[2]

Timeline and Next Steps

The August 2026 go-live date for Yotta’s new AI hub is notably soon—less than six months away from the announcement. This suggests that significant planning and groundwork have already occurred. The company will need to secure GPU allocations from Nvidia, manage complex logistics for deploying tens of thousands of processors, and ensure adequate power and cooling infrastructure at its facilities.

The IPO timeline of within 12 months creates an interesting dynamic: Yotta will likely be listing as a company that has just brought its flagship $2 billion infrastructure investment online, allowing investors to assess actual operational performance rather than relying purely on projections.

Looking Ahead

Yotta Data Services’ $2 billion AI hub represents more than just a single company’s expansion—it signals India’s determination to become a major player in global AI infrastructure. As the company moves toward its public debut, it will be closely watched by investors, policymakers, and technology leaders seeking to understand how emerging markets can compete in the AI compute race. The success of this venture could reshape the geography of AI infrastructure development for years to come.

Original source: CNBC Business – An Indian company is set to build a $2 billion AI hub with Nvidia’s GPUs and go public. Here’s what we know so far