(now NTT Global Data Centres and Cloud Infrastructure division),
Cyfuture India, Sify Digital Services, Vensysco Technologies, Locuz Enterprise Solutions, Yotta Data Services, and Ishan Infotech.
ET had reported on May 13 that the Centre has received bids proposing to offer a
total of 18,000 GPUs in the second round of the IndiaAI GPU tender and that it expects 15,000 GPUs to finally be offered.
Cyfuture has proposed to offer 1,184 GPUs, including Nvidia’s H100, L40S and A100 GPUs, AMD's MI300 and MI325 GPUs, and Intel's Gaudi 2 and Gaudi 3 GPUs. It has placed purchase orders for the same, people cited above said.
Vensysco has proposed to offer 2,300 GPUs: 2,000
Nvidia H100 GPUs, 100 AWS Trainium 1 GPUs, and 200 AWS Inferentia 2 GPUs.
Yotta has proposed to offer Nvidia Blackwell B200s.GPU models that NTT-Netmagic, Sify, Locuz and Ishan propose to offer are not known.
While Vensysco and Locuz are Amazon Web Services partners, Ishan is an Oracle partner and NTT-Netmagic is a leading Google Cloud partner in India.
Under the first round of the GPU tender concluded in February, IndiaAI Mission is offering 14,517 GPUs on subsidised rates to the country's startups, academics and research organisations.
ET had reported on April 26 that the Ministry of Electronics and Information Technology (MeitY) had assigned artificial intelligence (AI) projects, or workloads, to three companies selected to supply GPUs under the IndiaAI Mission.
The government, however, has not named the beneficiaries of the programme.
India gave the cabinet approval for the Rs 10,000-crore India AI Mission in March last year, with a target of creating an artificial intelligence (AI) compute infrastructure by procuring over 10,000 GPUs.
As part of the mission, the government is also incentivising the development of local language models built by academia and industry with investment capital and other support. The move is aimed at building up India’s AI prowess.
ET reported on May 16 that the IndiaAI Mission has received 506 foundation AI model proposals across three phases.
As part of the first phase of approvals, Sarvam.ai was selected to initiate the development of an indigenous foundational model. Sarvam will get access to 4,096 Nvidia H100 GPUs for six months from the IndiaAI Mission's common compute cluster to train its model.
