India Had a GPU Problem. Neysa Just Raised $1.2 Billion to Solve It.
In February 2026, Blackstone — the world's largest alternative asset manager, with $1.3 trillion in assets under management — made one of the largest bets in the history of Indian technology infrastructure. The target was not a consumer internet company, not a fintech, not a SaaS platform. It was a two-year-old AI infrastructure startup that most people outside the deep-tech investment community had never heard of.
The startup was Neysa. The deal was $1.2 billion. And the problem it was built to solve is one that determines whether India's AI ambition becomes a reality or remains dependent on foreign infrastructure.
Ganesh Mani, a Senior Managing Director at Blackstone Private Equity, estimated that India currently has fewer than 60,000 GPUs deployed and expects that figure to scale nearly 30 times to more than 2 million in the coming years. The gap between those two numbers — 60,000 GPUs today and 2 million required — is the infrastructure problem that Neysa was founded to close. Every enterprise that wants to train an AI model in India, every government body that needs sovereign compute for its AI applications, every global AI lab that wants India-resident infrastructure to serve its largest non-US user base: all of them run into the same ceiling. The compute is not there. Or it is in AWS and Azure data centres that are not in India, routing sensitive data through foreign jurisdictions in ways that regulated Indian industries increasingly cannot accept.
Neysa's answer to that problem just got $1.2 billion of capital behind it.
The Deal Structure — What $1.2 Billion Actually Means
The $1.2 billion capital raise comprises $600 million in equity capital from Blackstone and co-investors, with Neysa intending to secure an additional $600 million in debt financing. The equity investors alongside Blackstone include Teachers' Venture Growth, TVS Capital, 360 ONE Assets, and Nexus Venture Partners — which was already an existing investor from the Series A and re-invested in this round. Blackstone holds a majority stake following the transaction.
Prior to this round, Neysa had raised $20 million in seed funding in early 2024, followed by a $30 million Series A round in October 2024, backed by Nexus Venture Partners, NTT Venture Capital, Z47, and Anchorage Capital. The total pre-Blackstone capital was approximately $50 million. The Blackstone round is twenty-four times that amount.
The debt structure is the mechanism that makes the economics work. GPU infrastructure is capital-intensive in a way that is fundamentally different from software. Purchasing and deploying GPU clusters — the compute, networking, storage, and facility costs — requires the kind of balance sheet that institutional-grade debt financing provides. Blackstone's majority ownership means Neysa can access structured debt at institutional rates, and this capital cost advantage versus domestically-funded competitors creates a virtuous cycle: cheaper debt means more GPUs, which means lower per-unit cost, which means more competitive pricing, which means more market share.
The valuation that resulted from the transaction: approximately $1.4 billion, making Neysa India's 126th unicorn and the second to achieve that milestone in 2026 after Juspay.
Who Built Neysa — and Why His Background Explains the Company
Sharad Sanghi is not a first-time founder. He built Netmagic Solutions — one of India's pioneering managed IT services and data centre companies — and sold it to NTT Communications in 2012, a landmark transaction in the Indian infrastructure technology space. He spent years at NTT after the acquisition. He brought Nexus Venture Partners' backing from the Netmagic era with him when he founded Neysa — Nexus reinvested, reflecting a conviction based on watching Sanghi build something significant once and backing him to do it again.
Neysa was founded in 2023 with a specific and clearly stated founding thesis: India's AI ambition requires production-grade infrastructure built and operated at scale within India. Not access to foreign hyperscaler regions. Not workarounds for data residency requirements. Purpose-built, India-resident, sovereign GPU infrastructure.
Sanghi described the mandate at the time of the Blackstone announcement: "Neysa is focused on delivering the execution layer of sovereign compute, and AI research enablement and adoption in alignment with the goals of IndiaAI Mission. We seek to provide performance certainty and data assurance, enabling enterprises, hyperscalers, and global AI labs to deploy and scale reliable AI infrastructure in India. With Blackstone's experience in scaling critical infrastructure, we aim to help establish India as a globally relevant AI compute destination."
The phrase "execution layer of sovereign compute" is the precise description of what is missing in India's AI stack. The policy ambition exists — the IndiaAI Mission, the government's GPU procurement programmes, the National AI Strategy. The enterprise demand exists — every Indian bank, every healthcare provider, every insurance company that needs to run AI on sensitive data that cannot leave Indian jurisdiction. The gap is the infrastructure between the ambition and the application. Neysa is building that infrastructure.
What the Capital Is Building
The startup currently has about 1,200 GPUs live and plans to sharply scale that capacity, targeting deployments of more than 20,000 GPUs over time as customer demand accelerates. Sanghi told TechCrunch that the company expected to more than triple its GPU capacity within nine months of the announcement, based on customer conversations already at advanced stages.
The bulk of the capital will fund large-scale GPU clusters — compute, networking, and storage — while a smaller portion goes toward R&D and Neysa's software platforms for orchestration, observability, and security. The geographic footprint is expanding from its current Mumbai base across Delhi NCR, Bengaluru, and Chennai — the three cities where India's AI enterprise customer base is most concentrated.
The Telangana MoU for a ₹10,500 crore AI data centre cluster with NTT DATA marks the first multi-city anchor deployment outside Mumbai, creating the distributed compute geography that a nationally-serving AI infrastructure platform requires.
Neysa's product suite — branded Neysa Velocis — covers GPU-as-a-Service, AI Platform-as-a-Service, and Inference-as-a-Service. The software layer that Sanghi built before the Blackstone capital — integrating MLOps, monitoring, inference, and a marketplace — is the differentiation that positions Neysa above the commodity GPU pricing tier where Indian competitors like E2E Networks and NeevCloud compete primarily on price.
Sanghi has described the customer value proposition with specificity: "A lot of customers want handholding, and a lot of them want round-the-clock support with a 15-minute response and a couple of our resolutions. And so those are the kinds of things that we provide that some of the hyperscalers don't."
Why Blackstone — and What the Global Context Tells You
Blackstone is not a traditional technology venture investor. It is the world's largest alternative asset manager, and its conviction around AI infrastructure is not an India-specific bet. It is a global thesis being expressed in a specific Indian context.
Blackstone affiliates' key AI infrastructure investments include QTS, the world's largest data centre platform; AirTrunk, the leading data centre platform in the Asia Pacific region; CoreWeave, a specialised cloud infrastructure company; and Firmus, an Australian-based AI infrastructure platform. The Neysa investment adds the India layer to a global portfolio of AI infrastructure platforms — and it brings Neysa access to Blackstone's GPU procurement relationships, data centre construction expertise, and structured financing capabilities that no domestically-funded Indian competitor can access.
Amit Dixit, Head of Asia Private Equity at Blackstone, described the strategic framing: "Over the past two decades, we have been committed to building businesses that build India, and this investment brings that to life. It reinforces Blackstone's focus on backing the essential 'picks and shovels' of AI globally, including in India, a key market for Blackstone."
"Picks and shovels." That phrase — borrowed from the gold rush analogy that describes the companies that make money selling tools to the miners rather than mining themselves — is the most precise articulation of what AI infrastructure represents as an investment category. Every AI model that gets trained, every AI application that gets deployed, every AI inference that runs in production requires compute. The companies that provide that compute are the picks and shovels of the AI economy. Blackstone is buying the shovels. Neysa is the shovel manufacturer for India.
What India's AI Stack Now Has That It Did Not Before
The significance of the Neysa round extends beyond one company's capitalisation. It is a signal about what becomes possible in India's AI ecosystem when the infrastructure layer is properly funded.
For Indian enterprises in regulated industries — banking, insurance, healthcare — data residency requirements have been a structural barrier to adopting advanced AI. Routing sensitive customer data through US-based hyperscaler infrastructure for AI processing is not just technically suboptimal. In some regulatory contexts it is legally untenable. Neysa's India-resident, sovereign compute addresses that barrier directly.
For global AI labs — OpenAI, Anthropic, Google DeepMind — India is among their largest and fastest-growing user bases. The latency of serving Indian users from US or Singapore data centres is a real and measurable cost to product quality. India-resident GPU infrastructure, operated by a company with Blackstone's balance sheet behind it, offers those labs a credible option they did not previously have.
For Indian AI startups building foundation models — Sarvam AI, Krutrim, and others — the absence of affordable, accessible GPU capacity in India has been a constraint on the pace and ambition of what they can build. A 20,000-GPU Neysa infrastructure is the compute resource that changes the constraint equation.
The $1.2 billion is the largest AI infrastructure investment in India's history. But the number it is ultimately funding is not the $1.2 billion. It is the 2 million GPUs that Blackstone believes India will need — and the gap between 60,000 deployed today and that destination. Neysa is not the entire answer to that gap. It is the most credible beginning of one.
