India's $200 Billion AI Infrastructure Moment
India's data center power capacity sits at just over 1 gigawatt today. By 2030, projections put it above 8 GW — an 8x expansion driven by over $30 billion in direct capital investment and the most aggressive government incentive program any country has offered for AI infrastructure.
This isn't speculative. The commitments are signed and construction is underway.
The investment landscape
India's government has set a target of attracting over $200 billion in AI infrastructure investment by 2028. The major commitments so far:
- Amazon: Up to $35 billion by 2030
- Microsoft: $17.5 billion in data center investment
- Google: $15 billion for its first Indian AI hub
- Reliance Industries: $110 billion over seven years for AI data centers and edge compute
- Adani Group: $100 billion to expand from 2 GW to 5 GW of data center capacity
Why India, why now
Three structural factors are converging:
Policy: India's Union Budget 2026-27 introduced a tax holiday through 2047 for foreign cloud providers and a 15% safe harbour for domestic operators. Data centers have been granted infrastructure status, unlocking cheaper financing. This is the most favorable tax environment for AI compute anywhere in the world.
Power: India added 44.5 GW of renewable energy capacity in 2025 alone, reaching 254 GW total. The renewable build-out is directly enabling the data center expansion — green power is both cheaper and increasingly abundant. Electricity costs run 40-60% below U.S. data center markets.
Talent: India's semiconductor mission (ISM) has committed approximately $10 billion in incentives for chip fabrication and packaging. Combined with the country's existing engineering workforce, India is building specialized AI infrastructure capability at scale.
What this means for AI teams
For organizations running sustained AI workloads, the implications are practical:
Cost: The combination of tax exemptions, cheaper power, and competitive labor creates a structural TCO advantage for India-based compute. For multi-year training contracts or high-volume inference serving, the savings are measured in millions.
Capacity: The 8x expansion in data center capacity means supply is growing faster than demand. This is a buyer's market — pricing will remain competitive as providers compete for customers.
Options: With every major hyperscaler and multiple domestic providers building in India, AI teams have genuine choice in hardware, location, and contract terms.
Connectivity: India's submarine cable infrastructure and peering capacity is expanding in parallel with data center builds, reducing the latency concerns that historically limited India as a compute destination for global workloads.
Where NovaCore fits
We were early to this thesis. NovaCore deployed India's first NVIDIA Blackwell GPU cloud in Hyderabad in August 2025, before most of these macro investments were announced.
Our customers benefit from the ecosystem effects — better power infrastructure, more network capacity, and a deeper talent pool — while getting the focus and responsiveness of a dedicated GPU infrastructure provider rather than a hyperscaler's generic compute offering.
As India's AI infrastructure market matures, we're expanding our capacity and hardware offerings to match. Get in touch to discuss how NovaCore's infrastructure can serve your workloads.