2 days ago
India Builds Sovereign AI Compute Pool Amid Rising GPU Costs
India wants to create its own pool of powerful computers for artificial intelligence.
The government is working with C-DAC to add computing power equal to about 3,000 GPUs.
It is also finalizing a deal with the UAE for a supercomputer equal to about 10,000 GPUs.
The UAE-supported system would be owned by the Indian government.
India’s AI Mission has access to about 30,000 GPUs, although 45,000 were promised.
Some private companies have struggled to provide all the promised machines because equipment has become more expensive.
The government has not yet decided how much users will pay for the computing power.
Analysts are asking whether India will use the newest and most efficient chips.
The Centre is developing a government-controlled AI computing pool with about 10,000 GPUs of capacity.
C-DAC is working to install 3,000 GPUs, while an UAE-backed 8-exaflop supercomputer could provide capacity equivalent to about 10,000 GPUs.
The supercomputer is being developed by G42 and Cerebras with C-DAC and Mohamed bin Zayed University of AI.
The India AI Mission currently has access to about 30,000 GPUs against commitments of 45,000, as equipment costs rise.
Analysts want clarity on chip generations, intended uses, and whether the mission’s ₹10,372 crore budget will be increased.
- Who
- The Indian government, C-DAC, the UAE, G42, Cerebras, and Mohamed bin Zayed University of AI are involved.
- What
- The Centre is building a government-controlled AI computing pool and planning additional computing capacity.
- Where
- The computing facilities will be set up in India.
- When
- The plans were reported on October 2, 2026; installation and agreements are still being finalized.
- Why
- The government wants to expand sovereign AI capacity as private providers face rising equipment costs and delivery shortfalls.
Government Expansion Plan
Analyst Concerns
Building sovereign capacity
Government Expansion Plan
The government is combining C-DAC capacity with a UAE-supported supercomputer and plans to seek further bids to create a government-controlled AI compute pool.
Analyst Concerns
Analysts say the government needs to provide more clarity about the ecosystem it is building, including the specific accelerator chips and intended users.
Chip quality and efficiency
Government Expansion Plan
The government is moving ahead with plans to install additional computing capacity, with pricing and procurement details still under development.
Analyst Concerns
Neil Shah warned that relying mainly on chips that are three to four years old could reduce power and performance efficiency.
Funding adequacy
Government Expansion Plan
The government has set a ₹10,372 crore budget for the India AI Mission over five years and is reviewing subsidies based on startup needs.
Analyst Concerns
Analysts question whether the budget will need to be increased to keep pace with rising costs and secure advanced chips.
Key facts
- Planned C-DAC capacity
- 3,000 GPUs of computing capacity
- UAE-supported system
- An 8-exaflop AI supercomputer, equivalent to about 10,000 GPUs of capacity
- Current mission access
- About 30,000 GPUs
- Committed mission capacity
- 45,000 GPUs
- India AI Mission budget
- ₹10,372 crore over five years
- Ownership
- The UAE-supported supercomputer is intended to be owned by the Indian government as a UAE government gift
- Pricing status
- GPU pricing and subsidy levels have not yet been decided
Quotes
A senior government official
Government official involved in the AI computing agreements
“We are working with C-DAC on a 3,000 worth of capacity to be installed and then we are also working on the other G42 termsheet which was signed in the presence of the Prime Minister. We are doing the final agreements to get further compute. That will be worth of about 10,000 GPUs worth of compute.”
thehindubusinessline.com
“We will have to see...we will have to adjust the subsidies according to what the startups get. It will take at least three-four months and then we are going to float bids... It is still in RFP stage.”
thehindubusinessline.com
Sources
Related news
AM Intelligence Orders 20,000 Nvidia GPUs in $4 Billion Deal
Goyal, Microsoft Discuss Data Centres and India AI Mission
NTA Uses Sovereign AI for NEET-UG Translation After Paper Leak
TCS HyperVault Plans ₹70,000-Crore AI Campus in Hyderabad
Nvidia Price Hike Puts India’s AI Data-Centre Build-Out Under Pressure
India’s AI Enablers Gain as Data-Centre Buildout Accelerates
IIT Delhi Abu Dhabi Campus Expands
