India's Yotta Crosses the Pacific: Why Nidar's Florida MOU Is a Cross-Border AI-Power Bet

Tachyon9 Corporation said it has signed a binding 15-year Memorandum of Understanding with Nidar Infrastructure Limited, the parent and majority shareholder of India's Yotta Data Services, to anchor the first 100-megawatt phase of its planned Nakota AI Data Campus in the United States. Tachyon9 said the commitment is expected to contribute roughly $2.34 billion in revenue over the term, tied to that initial phase, the first slice of a campus management has designed to scale to as much as 1 gigawatt of capacity. The deal also underpins Tachyon9's proposed combination with Nixxy, Inc. (NASDAQ: NIXX), the public vehicle through which the partners intend to take the platform to market.

What the agreement actually commits

Under the MOU, Tachyon9 said Nidar's U.S. affiliate is expected to become the anchor customer for the first 100 MW at Nakota while also participating as an economic partner. At full utilization of that opening phase, the company said the arrangement contemplates roughly $156 million in annual contracted infrastructure revenue; run across the 15-year term, that yearly figure is what builds to the $2.34 billion the company attributes to the first 100 MW. The scope matters: the $2.34 billion is the multi-year total for the first phase, not an annual run-rate and not a figure for the full 1 GW target. The wider campus is described as up to ten times the size of this first tranche, but Tachyon9 has not extended the revenue math to that ceiling, and neither should anyone reading the release.

For its part, Tachyon9 said it is contributing roughly $64 million in equipment and land-option rights, along with a signed letter of intent covering the entire 1 GW development. Both companies framed the news as a strategic binding MOU and a proposed combination, a framework, not a closed transaction or contracted full-campus revenue.

The sector backdrop

This is one story in a much larger build-out. Hyperscale demand for AI compute has turned power and cooling into the binding constraint of the data-center cycle, and operators are racing to lock in long-dated capacity contracts before the silicon arrives. A 15-year anchor commitment of the kind Tachyon9 describes is the contractual shape that build-out tends to take: a developer secures recurring, long-term revenue to underwrite power infrastructure that can take years to energize. The demand environment is real and sector-wide; what remains to be proven, here as elsewhere, is execution against a single early-stage site.

The cross-border read-through

The more distinctive angle is geographic: the partner stepping into a Florida-anchored project carries credentials built in a different market. Tachyon9 described Yotta as India's largest operator of Nvidia GPU compute infrastructure, holding an estimated 60-70% of the country's deployed GPU capacity and running campuses in Navi Mumbai, Gujarat and Greater Noida, with a fourth planned in Telangana. In February 2026, Yotta announced a $2 billion-plus investment to deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs at Greater Noida, paired with a four-year NVIDIA DGX Cloud engagement the company valued at over $1 billion and an allocation of more than 10,000 GPUs to the Government of India's IndiaAI Mission. Those are Yotta's stated figures, drawn from the release.

That footprint now reaches across borders, a read-through for the broader cycle: capacity, capital and customer demand are increasingly arranged across regions rather than within them. Darshan Hiranandani, Chairman of Nidar and Co-Founder of Yotta, cast the move in those terms, saying the Nakota campus "brings this proven model to the United States at exactly the moment when America's AI industry needs it most."

Who else this touches

Sitting alongside the operating story is a capital-markets one. Tachyon9 said Yotta is pursuing pre-IPO financing and targeting a public listing at a valuation of roughly $4-6 billion, with plans to raise $600-900 million in growth capital, which the company offered as a gauge of investor appetite for AI-infrastructure assets. For U.S. public-market investors, the accessible piece is Nixxy, the listed company Tachyon9 proposes to combine with. Shahal Khan, Tachyon9's Chairman and CEO, cast the agreement as "strategic validation from one of the world's leading AI infrastructure operators." That is management's characterization; the read for everyone else is more measured.

A balanced close

Stripped to its load-bearing facts, this is a binding MOU and a proposed combination, not a completed merger or contracted full-campus revenue, and the headline projections are the companies' own. Within those limits, it is a clean illustration of where the AI-infrastructure cycle is heading: toward long-dated power contracts and cross-border partnerships that braid Asian operating expertise into the U.S. build-out. Whether the first 100 MW energizes on schedule, and whether the path to 1 GW follows, is what will tell over the coming years. For now the signal is the direction of travel, not the destination.

Disclosure: NewsHK is an independent financial-news publication. This article is editorial content and not investment advice. NewsHK holds no position in the securities mentioned and was not compensated for this coverage. Figures are drawn from the company's press release.