Thursday, 30, January 2025: Union IT minister Ashwini Vaishnaw at the Utkarsh Odisha Conclave disclosed that India is planning to develop its own generative AI model. This initiative will leverage the 18,693 GPUs secured by the India AI Compute Facility to build a large language model (LLM).

This initiative will be a significant step for India in the field of AI. It will enable the country to develop AI models tailored to its specific needs which can be leveraged to address various emerging challenges. Additionally, having its own generative AI model will minimise India’s reliance on foreign AI models.

Vaishnaw announced, “I am very happy to say that, against 10,000 GPUs, we have actually empanelled 18,693 GPUs,”. This includes 12,896 Nvidia H100s and 1,480 Nvidia H200s, some of the most powerful GPUs available. “Overall, the net high-end GPUs we have is 15,000,” he added, noting that even a model like DeepSeek was trained on just 2,000 GPUs. Roughly 10,000 of these GPUs are available today.

“The second big mission of the IndiaAI mission was to develop an AI model,”. “We have now created the framework which will be launched today. We are calling for proposals to develop our own foundational models, where the Indian context, the Indian languages, and the culture of our country are considered, where the biases can be removed. Where the datasets, for our country, for our citizens – that process will start today.”- he added.

The development of the generative AI model is expected to be completed in four to eight months. Once it is completed, not only it is going to be a valuable tool for researchers, startups, and several institutions in India but also help to promote the development of AI in India.

Source: https://economictimes.indiatimes.com/tech/artificial-intelligence/ashwini-vaishnaw-says-india-to-develop-own-generative-ai-model-report/articleshow/117724681.cms?from=mdr