IndiaAI Mission: Jio Platforms, Tata Submit Bids To Offer AI Compute

The electronics and IT ministry (MeitY) said that 19 service providers have submitted bids to provide artificial intelligence (AI) compute and cloud services under the INR 10,738 Cr IndiaAI Mission. 

The entities that participated in the bidding round included cloud service providers, managed service providers, micro, small and medium enterprises (MSMEs), and data centre service providers, including Jio Platforms, Tata Communications, Sify Digital Services, Yotta, CMS Computers India, among others. 

CloudThat Technologies, Cyfuture India, I2k2 Networks, Ishan Infotech, Orient Technologies Limited, NxtGen Datacenter and Cloud Technologies, Path Infotech, Shezar Web Technologies, Unicloud Labs and Vensysco Technologies were the other bidders.

The selected companies will be empanelled as agencies to provide AI compute services to startups, researchers, students, and academicians. Under the IndiaAI compute pillar, the Centre plans to build infrastructure comprising 10,000 graphics processing units (GPUs) in partnership with private players. 

In a statement, MeitY said that the pre-bid meeting in August 2024 saw participation from 50 service providers. Subsequently, November 28 was set as the deadline for the bid submission process and the bids were opened on December 2. 

Now, a technical evaluation committee will evaluate the submitted bids based on the eligibility and technical criteria specified in the request for empanelment (RFE) document. Afterwards, the eligible bidders will be directed to present their proposed solutions before the panel. Only then will the “technically qualified bidders” be allowed to participate in the commercial bids. 

The ministry also said that empanelment will be valid for a period of 36 months. There will, however, be an option to extend the timeline for an additional 12 months based on mutually agreed terms, it added. 

The INR 10,738 Cr IndiaAI Mission was approved by the union cabinet earlier this year. It focusses on seven areas – computing, foundational models, datasets, applications, skill development, startups, and safe AI. It aims to foster the homegrown GenAI ecosystem via a public-private partnership (PPP) model. 

In May this year, MeitY secretary S Krishnan said that domestic companies will be preferred when it comes to adding GPU power approved under the Mission. At the time, Krishnan had said that AI compute capacity is expected to be made available to startups and researchers in the next 18-24 months, with the first set of use cases expected by the end of the financial year 2024-25 (FY25).

Afterwards in September, Krishnan also said that the government might look at viability gap funding (VGF) to create more “compute capacity” once the existing 10,000 GPU target has been achieved.