IndiaAI jumps on artificial intelligence bandwagon

Yotta, backed by the Hiranandani Group, Jio Platforms, Tata Communications, E2E Networks, CMS Computers, Ctrls Datacenters, Locuz Enterprise Solutions, NxtGen Datacenter, Orient Technologies and Vensysco Technologies are helping build the state AI tool.

author-image
The Squirrels Bureau
New Update
Listen to this article
0.75x 1x 1.5x
00:00 / 00:00

Following the recent launch of the affordable foundational model DeepSeek by a burgeoning Chinese artificial intelligence (AI) laboratory, the Indian government has announced its intention to develop a domestic large language model as part of the Rs 10,370 crore IndiaAI Mission, as stated by IT Minister Ashwini Vaishnaw on January 30.

Additionally, the government has identified ten companies tasked with supplying a total of 18,693 graphics processing units (GPUs), which are essential high-performance chips for the creation of machine learning tools necessary for the foundational model's development. The selected companies include Yotta, backed by the Hiranandani Group, Jio Platforms, Tata Communications, E2E Networks, CMS Computers, Ctrls Datacenters, Locuz Enterprise Solutions, NxtGen Datacenter, Orient Technologies, and Vensysco Technologies. Notably, Yotta is set to provide nearly half of the total GPUs, committing to deliver 9,216 units.

“Over the last 1.5 years, our teams have been working closely with startups, researchers, professors etc.  Today, we are calling for proposals to develop our foundational model. The model will take care of the Indian context, languages and culture, are devoid of biases,” Vaishnaw told reporters.

The minister announced that the government is currently engaging with at least six developers to create the foundational model, which is expected to take approximately 4 to 8 months to complete. He stated, “We will have a world-class foundational model in the next few months.” However, he did not disclose the estimated cost of developing the model or the identities of the companies involved in discussions with the government.

Government's preparedness and competence

Regarding the procurement of computing power, Vaishnaw indicated that out of the 18,693 GPUs approved for empanelment, around 10,000 GPUs are ready for installation already.

In the coming days, the government plans to introduce a shared computing facility that will allow startups and researchers to access computing resources. The fee for utilizing high-performance GPUs will be Rs 150 per hour, while the cost for lower-performance GPUs will be Rs 115.85 per hour. To facilitate access to these services, the government will provide a 40% subsidy to end users on the total charges.

“Globally, GPU access costs $2.5-$3 per hour. We are making it available, after the subsidy for around $1 per hour,” Vaishnaw said.

The government, as part of the IndiaAI Mission, has identified 18 application-level AI solutions for initial funding. Vaishnaw indicated that these applications are concentrated on agriculture, learning disabilities, and climate change.

China artificial intelligence
Advertisment