Union Minister for Electronics and Information Technology, Ashwini Vaishnaw, has announced that India is developing its own AI foundation model, set to be ready within 10 months. Backed by a robust infrastructure, India’s AI mission is gaining momentum, with the country surpassing its initial target of securing 10,000 GPUs and now boasting an impressive 18,600 GPUs. This advanced computing power is critical for training the AI model and positions India as a strong contender in the global AI race.

India’s AI Infrastructure: A Game-Changer

India’s AI ambitions are supported by cutting-edge infrastructure, including high-performance NVIDIA H100 and H200 GPUs, as well as MI325 models. To put this into perspective:

• DeepSeek AI was trained on 2,500 GPUs.
• ChatGPT was trained on 25,000 GPUs.
• India now has over 15,000 high-end GPUs, with more being added to the pool.

This infrastructure will play a pivotal role in developing AI models that can rival global competitors. Vaishnaw emphasized that the availability of such resources will accelerate India’s progress in AI innovation.

Democratizing AI Development

To further support the AI ecosystem, the Indian government has introduced a common computing facility equipped with 18,000 GPUs. This facility will be accessible to AI startups, developers, and researchers, enabling them to leverage advanced computing resources without incurring prohibitive costs. Currently, 10,000 GPUs are operational, with the remaining set to come online soon.

Vaishnaw highlighted the importance of this initiative in democratizing AI development, allowing smaller players to contribute to India’s AI mission. This move is expected to foster innovation and collaboration across the AI community.

India’s AI Model: Designed for Diversity

India’s AI foundation model is currently under development, with six major developers working on the project. The model is being designed to handle India’s vast linguistic and cultural diversity, making it uniquely suited to address local challenges while competing on a global scale.

The government expects the first versions of the model to be ready within 4 to 10 months. Vaishnaw expressed confidence in the project, citing advancements in algorithmic efficiency that will enable India to produce a world-class AI model in a shorter timeframe.

Conclusion

With a strong infrastructure, a focus on democratizing resources, and a commitment to addressing India’s unique needs, the country is well on its way to becoming a global leader in AI. The development of its own foundation model, supported by 18,600 GPUs, marks a significant milestone in India’s AI journey. As Ashwini Vaishnaw noted, this initiative will not only boost innovation but also position India as a key player in the global AI landscape.