India’s AI moment is no longer just about models and applications. It is increasingly about the hard infrastructure underneath — energy-hungry data centers, high‑performance compute, sovereign cloud stacks, and large‑scale chip partnerships that will determine how far and how fast the country can move in the global AI race.
Over the past week, a series of announcements from global and Indian players have signalled the start of an AI infrastructure supercycle in India, spanning tens of billions of dollars in planned investments.
Adani’s 100 billion bet on AI-ready data centers
Adani Group has laid out one of the most ambitious infrastructure commitments yet, with plans to invest 100 billion dollars by 2035 in renewable‑powered, AI‑ready data centers across India. The group intends to scale from its existing data center capacity to as much as 5 GW of AI compute, positioning this as part of a five‑layer “AI stack” focused on India’s technological sovereignty.
The company expects its outlay to catalyse as much as 150 billion dollars of additional investment in adjacent areas such as server manufacturing and advanced electrical infrastructure, potentially helping create a 250 billion dollar AI infrastructure ecosystem in the country over the next decade. The strategy also links tightly to India’s renewable energy ambitions, with AI facilities planned to be powered by large‑scale green energy assets.
Microsoft’s India-first cloud and AI build-out
Microsoft is emerging as another major pillar of this build‑out. The company has committed 17.5 billion dollars to India over four years, its single largest bet in Asia, spanning cloud and AI infrastructure, skilling, and ongoing operations. A significant part of this investment is going into new data center regions and capacity, with Hyderabad emerging as a key hub.
Crucially, Microsoft is framing this expansion through the lens of data and AI sovereignty. Indian customers and the government increasingly want data processed and stored within the country, under local regulatory frameworks, and the company has positioned its infrastructure roadmap to meet these expectations. This aligns with sectors such as banking, healthcare, and public services, where in‑country residency for sensitive data is non‑negotiable.
Microsoft’s wider Global South strategy — with up to 50 billion dollars earmarked for AI investments across emerging markets — further underscores India’s centrality as both a deployment base and an innovation hub.
L&T–NVIDIA’s gigawatt-scale AI factory vision
If Adani and Microsoft are signalling scale through dollars and regions, Larsen & Toubro (L&T) and NVIDIA are doing it through architecture. The two companies have announced a plan under the IndiaAI Mission to build what is being described as India’s largest gigawatt‑scale AI factory.
The proposed venture combines L&T’s strengths in engineering and infrastructure execution with NVIDIA’s full AI stack — GPUs, high‑speed networking, accelerated storage, and the NVIDIA AI Enterprise software platform. The goal is to create sovereign, scalable infrastructure that can host large‑scale AI workloads across manufacturing, energy, financial services, healthcare, and public services within India’s borders.
Designed as “sovereign by design”, the fabric is intended to serve domestic requirements while remaining interoperable with global cloud ecosystems, giving enterprises and hyperscalers the option to deploy high‑density AI capacity from India as a strategic hub.
Yotta and the rise of AI hubs
Adding to this wave, Yotta is building a two‑billion‑dollar AI data center in India powered by NVIDIA’s Blackwell‑generation chips, positioning it as one of Asia’s largest AI hubs. Facilities of this scale are being designed specifically for large‑language models, generative AI, and high‑performance enterprise workloads, rather than traditional colocation alone.
Together, these moves indicate that India is not just adding incremental capacity, but seeding specialised AI infrastructure clusters that can serve both domestic demand and global workloads.
NVIDIA’s role in the India AI Mission
NVIDIA’s own articulation of the India AI Mission ties these threads together — infrastructure, models, and ecosystem. The company has outlined plans to support national‑scale AI infrastructure, collaborate on foundation models tuned to Indian languages and use cases, and enable industry and academic partners through reference architectures and software platforms.
This approach dovetails with India’s push for digital public infrastructure in AI — sometimes described as moving from India Stack to an India “AI stack” — where compute, data, and models are treated as shared capabilities that multiple players can build on.
A new phase of AI competition
Viewed together, Adani’s 100 billion dollar programme, Microsoft’s 17.5 billion dollar India commitment, the L&T–NVIDIA AI factory, Yotta’s Blackwell‑powered hub, and NVIDIA’s broader India AI Mission represent more than isolated announcements. They signal the start of an infrastructure race that will determine who controls the pipes, platforms, and performance characteristics of India’s AI future.
For India, the opportunity is twofold: to meet exploding domestic demand for AI compute, and to position itself as a strategic AI and cloud region for the Global South and beyond. The policy choices made around sovereignty, openness, and sustainability will now shape how inclusive and resilient this AI infrastructure supercycle becomes.
