10,000 GPUs Up for Grabs as C-DAC Opens IndiaAI Bids

C-DAC opens technical bids today, 30 Oct 2026, for up to 10,000 AI compute units under IndiaAI Mission. What the tender asks and who qualifies.

Informeia News Desk
7 Min Read

India’s race to put serious computing power in the hands of its own startups, researchers and students takes a concrete step forward today. The Centre for Development of Advanced Computing, or C-DAC, opens the technical bids at 4 pm on Friday, 30 October 2026, for a tender that could add up to 10,000 AI compute units to the IndiaAI Mission’s shared pool.

The window for submitting proposals closed at 3:30 pm on Thursday, 29 October, on the Central Public Procurement Portal. What opens today is not a price auction. It is the technical qualification round, the stage where C-DAC checks which providers are actually capable of delivering before money is discussed at all.

For anyone building AI in India, this matters in a very practical way. Training modern AI models needs GPUs in large numbers, and renting them at market rates can be ruinously expensive for a small team. The IndiaAI Mission buys compute in bulk and offers it to eligible users at subsidised rates, and every new tranche widens the door a little further.

What the tender actually asks for

The Request for Empanelment, numbered C-DAC/GPU/01/2026 and published on 7 October, invites AI service providers to supply AI compute units for five years. The term is deliberately broad. An AI compute unit can be a GPU, a TPU or a similar accelerator, and the tender names high-end hardware such as NVIDIA’s H100 and H200 chips, along with next-generation equivalents, as the kind of kit it expects.

The scale is organised in clusters. Providers can offer capacity in blocks of 4,000, 1,000 or 512 units, with each cluster housed in a single data centre in India. C-DAC has also left room to scale beyond 10,000 units if demand from ministries, academia and industry justifies it. The second stage of the process will run in multiple rounds, so today’s opening is the start of the selection, not the finish line.

To get through Stage 1, a bidder needs at least 60 marks out of 100 on technical qualification, with no price bids involved. Successful bids must stay valid for 180 days.

Who can bid, and why the bar is high

The eligibility conditions point to a field of large, well-capitalised players. Bidders need an average annual turnover of at least Rs 120 crore and a Tier-III data centre in India. They must either already have 512 compute units installed or commit to deploying them in India within six months, backed by a bank guarantee.

That bar is not accidental. In earlier rounds, some providers could not deliver pledged units as equipment costs rose sharply, so C-DAC wants this tranche to turn into working machines, not paper commitments.

Earlier rounds featured Jio Platforms, Tata Communications, Yotta Data Services and firms such as CtrlS, E2E Networks and NxtGen. Under the Rs 10,372 crore mission, 14 companies are already empanelled and more than 38,000 GPUs onboarded. Add 10,000 more, and the national pool moves past 50,000.

Why 10,000 more GPUs, and why now

Union Electronics and IT Minister Ashwini Vaishnaw has linked the expansion to India’s wider AI push. At the World Bank’s World Development Report 2026 launch in New Delhi on 8 October, he put demand at 10,000 GPUs and backed empanelment over one-off tenders. On 11 October, he announced 10,000 more GPUs for the mission’s shared pool.

The timing sits alongside a second announcement from the same week: a consultation paper on AI regulation, expected within a month of 8 October, covering safety, AI agents and deepfakes. Rules and compute are moving together.

What happens next

  • Today, 30 October, 4 pm: C-DAC opens the technical proposals submitted by 29 October.
  • Technical scoring: Bidders need 60 out of 100 marks to be empanelled, with no price bids at this stage.
  • Later rounds: Stage 2 proceeds in multiple rounds, with capacity drawn down as demand arrives from eligible users.
  • Delivery: Empanelled providers must deliver end-to-end AI compute services over five years, with new capacity in Indian data centres.

For startups, the practical takeaway is simple: watch the empanelment outcome, because that decides which providers will offer subsidised compute and at what price. In earlier rounds, the average lowest price discovered was nearly Rs 115 per GPU hour, with high-end H100 capacity around Rs 140 per GPU hour.

Frequently asked questions

What is opening today? C-DAC opens technical bids at 4 pm on 30 October 2026 for its tender to empanel AI service providers under the IndiaAI Mission.

How many AI compute units are involved? The tender covers up to 10,000 AI compute units, including GPUs, TPUs and similar accelerators, with room to scale further if demand warrants it.

Who can use this compute? Ministries, departments, academia, research institutions, startups and industry can access the mission’s shared pool at subsidised rates once capacity is empanelled.

What do bidders need to qualify? At least 60 out of 100 marks in technical qualification, average annual turnover of at least Rs 120 crore, a Tier-III data centre in India, and either 512 installed units or a guaranteed deployment within six months.

When will the new GPUs be available? Availability follows technical qualification and the later Stage 2 rounds, so no single date has been fixed yet. Providers will deliver services over a five-year period.

Today’s bid opening moves this expansion from announcement to execution. For the country’s AI builders, that is the news that counts.

Ask about this articleInstant answers from this articleAI
AI-generated answers. Ask about the article, or anything else.
Share This Article
Informeia News Desk is the editorial team behind Informeia, covering breaking news across business, economy, politics, sports, entertainment, technology, health, education and jobs, with a focus on fast, accurate and reader-friendly reporting.
Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Choose language