India is scaling its national AI compute infrastructure faster than almost any other country. At the India AI Impact Summit 2026, Union IT Minister Ashwini Vaishnaw announced that orders for 20,000 new GPUs will be placed within a week, with deployment expected over the next six months — taking India’s total sovereign compute base from 38,000 to over 58,000 high-end GPUs under AI Mission 2.0.

India AI Mission 2.0: Key Numbers at a Glance

What makes India’s compute approach genuinely different from most countries: access is democratised. These GPUs are available to startups, researchers, and students at approximately ₹65 per hour — roughly one-third of the global average cloud compute rate.

Detail	Figure
Existing GPU Base	38,000 high-end GPUs (IndiaAI Mission)
New GPUs Being Added	20,000 — orders placed within 1 week
Total After Expansion	58,000+ GPUs
Deployment Timeline	Within 6 months of order
Access Cost	~₹65/hour — ~⅓ of global average
Expected AI Investment (2 years)	$200B across five layers of AI stack
VC Commitment (deep-tech)	$70M confirmed; more expected
Total Investment Projection	Up to $400B+ across AI stack layers
Deep-tech Startups Expected	50+ in coming years
Countries in AI Policy Dialogue	~30 nations
Energy Efficiency Target	35% reduction in AI infra energy use
Clean Energy Share	51% of India’s power generation capacity
Tech Sector Revenue (2026)	$280B projected
AI GDP Contribution by 2035	$1.7 trillion

What AI Mission 2.0 Means Beyond the GPU Count

The original IndiaAI Mission was about building the foundation — getting GPUs onboarded and making them accessible. Mission 2.0 shifts focus to R&D, innovation, and diffusion — turning raw compute into competitive IP and globally relevant AI products.

Vaishnaw described this as the “fifth industrial revolution” and framed India’s approach as structurally different from Western models. In most countries, AI infrastructure is controlled by a handful of corporations — Meta, Google, Microsoft, Amazon. India’s model puts sovereign compute in a common pool, accessible at subsidised rates to anyone in the ecosystem.

The semiconductor angle is equally significant. Semiconductor 2.0 will focus on chip design rather than fabrication — a pragmatic choice given the capital intensity of fabs, and one that plays to India’s existing engineering talent base. At least 50 deep-tech startups are expected to emerge from this focus in the coming years.

On AI governance, Vaishnaw confirmed India is in dialogue with approximately 30 countries on policy and regulatory frameworks, pushing for a “techno-legal” approach — and called for much stronger global regulation on deepfakes, particularly to protect children.

For the official government statement, visit the Press Information Bureau release.

Check our India AI Mission coverage on TechnoSports and India AI Impact Summit 2026 full roundup.

FAQs

How many GPUs will India have after the AI Mission 2.0 expansion?

India will go from 38,000 to over 58,000 high-end GPUs, with orders for the new 20,000 units placed within a week of the February 17 announcement and deployment expected within six months.

How much does it cost to access India’s sovereign AI compute infrastructure?

Approximately ₹65 per hour — around one-third of the global average cloud GPU rate — making it accessible to startups, researchers, and students across India.