The Central government’s ambitious India AI mission will finally take off as AI compute capabilities will be available by the mid-year 2025, a senior government official said on Thursday.
In August, the government had invited bids for the empanelment of entities for providing artificial intelligence services on cloud under the Rs 10,372-crore India AI mission. The process of empanelling of companies is expected to complete by the end of 2024.
“After the tender was floated, we have heard of various companies working towards investing in AI compute infrastructure so that they get empanelled. We hope to finalise this process by December. And in the next six months or so, we will see availability of compute within India,” said Abhishek Singh, Additional Secretary, Ministry of Electronics and Information Technology (MeitY).
Singh was speaking at the Nvidia AI Summit in Mumbai.
The empanelled agencies include sites and firms such as data centres, startups and cloud service providers.
These entities will have to provide access to high-speed computing AI infrastructure like graphics processing unit (GPUs), accelerators, Tensor Processing Units (TPUs), storage to academia, startups, researchers, government bodies among others at the lowest rate that will be discovered through the bidding process.
More From This Section
Meanwhile, Singh added that the country needs more investment when it comes to AI infrastructure.
“Infrastructure is one area in which we need to invest more. We need to work to ensure talented engineers have the compute infrastructure available to play with and come up with solutions solving societal problems at population scale,” he said.
The government is spending about Rs 5,000 crore of the Rs 10,372 crore fund under the India AI Mission to procure GPUs and provide subsidised compute capacity to Indian startups.
Under the India AI mission, supercomputing, comprising over 10,000 GPUs, will be made available to various stakeholders for creating an AI ecosystem.
Singh elucidated that the use of AI across digital public infrastructure such as payments and identity, among others, would bring in quantifiable efficiency.
“Using AI, we can make DPI 50-60 pc more efficient. That would give us the power of AI over DPI,” he added.