“L&T prefers to go as one team rather than compete. E2E will front or prime this for them. E2E will be the primary bidder. L&T is a consortium partner for them,” a source told ET.
L&T and E2E did not respond to ET’s request for comment.
The last date for filing the bids is November 28 (Thursday).
The tender is part of the government’s initiative to set up public AI cloud infrastructure with a capacity of at least 10,000 GPUs through a public-private partnership under the IndiaAI Mission. November 28 marks the third extension of the deadline to submit bids, which was originally set for September 6 and later pushed to October 16 and then to November 12.
While E2E procures the GPUs, L&T will provide the data centre infrastructure to house them. This consensus comes days after L&T on November 5 announced that it will acquire a 21 per cent stake in E2E, in an all-cash transaction of Rs 1,406 crore.
Discover the stories of your interest
The strategic investment, L&T executives had said, helps expand the company’s Cloud service offering as part of the data centre business. L&T currently owns two data centres – Panvel (2MW) and Chennai (12 MW), which it offers to third-party customers.E2E has co-located servers with Netmagic in Greater Noida with a capacity of 3.2MW housing the majority of its GPU infrastructure. Two smaller facilities in Mumbai and Noida house its CPUs.
As per the bid document, primary and secondary partners can pool resources to meet the eligibility criteria. Early September, Seema Ambastha, chief executive, data center business, L&T, had told ET, over email, “We meet all the pre-qualification and technical qualification criteria.”
Tarun Dua, chief executive of E2E Networks, had told ET late September, “The relaxations on revenue qualifications have allowed us to qualify directly, and we deeply appreciate this flexibility.”
NSE-listed E2E Networks late October had announced that it is the first company to bring NVIDIA H200 Tensor Core GPUs to the Indian market which delivers up to 1.9X higher inference performance compared with NVIDIA H100 Tensor Core GPUs.
This is expected to enable businesses to tackle more complex AI models and drive innovation across startups, MSMEs and large businesses alike.
The company had raised Rs. 420 crores from the market for procurement of the GPUs. It has so far received 256 H200 GPUs with more GPUs being procured soon.
E2E Cloud’s flagship product, TIR – an AI development studio – will be the first in India to feature H200 GPUs, giving developers access to cutting-edge infrastructure. This will allow developers to train foundational AI models. The company anticipates that the H200 GPU will become a driver for AI training in India, specifically for the training and inference of large language models (LLMs) and large vision models (LVMs).
E2E Cloud is used by startups, SMEs and large businesses from India, the Middle East, the Asia Pacific, and the US.
Other bidders for the tender include Jio Platforms, Amazon Web Services, NTT Data, and Yotta, among as many as 30 companies that include data centres and hyperscalers.
