GoI to invest at least Rs 5,000 Crore on high-end NVIDIA AI GPUs and chips for IndiaAI Mission

The Government of India (GoI) is gearing up to invest around ₹5,000 crore from the total ₹10,372 crore allocated for the India AI Mission, into GPUs, with speculations coming up that suggest a major bulk of this fund will go to NVIDIA.

This significant portion of the funding from India AI mission will be used to procure advanced Graphics Processing Units (GPUs) and provide subsidized compute capacity to Indian startups, as revealed by a senior official from the Ministry of Electronics and Information Technology (MeitY) during the Global IndiaAI Summit held in New Delhi.

GPUs are highly specialized chips that excel at accelerating artificial intelligence (AI) and machine learning tasks by enabling faster data processing. These chips, produced by leading companies such as Nvidia, Intel, and AMD, are critical components in building powerful and efficient AI systems.

By investing in these GPUs, the Indian government aims to boost the computational capabilities available to startups and other enterprises involved in AI research and development.

Abhishek Singh, the Additional Secretary of MeitY, explained that the government has earmarked nearly ₹5,000 crore to acquire over 10,000 GPUs. This initiative is part of the broader mission to enhance India’s AI infrastructure.

He noted that the government is in the process of issuing tenders for these procurements, but emphasized that the approach would not involve directly purchasing and managing the GPUs. Instead, the government seeks to encourage private sector investment and will subsidize part of the cost of accessing these compute resources.

Singh elaborated that the aim is not to build a government-owned, publicly-run infrastructure but rather to support private investments while making computing capacity more affordable. This subsidy will lower the costs for those who need computational power, such as startups developing AI models, conducting research, or working on complex algorithms.

By doing so, the government hopes to reduce the financial burden currently faced by these innovators and researchers.

One of the key aspects of this approach is allowing the end users to decide which GPUs best meet their specific needs. Whether a startup requires the H100, Gaudi 2, or any other specific model, the choice will be theirs.

The government will not impose decisions on which GPUs to use, recognising that startups and other enterprises are best positioned to determine the most suitable infrastructure for their AI projects. This user-centric procurement model ensures flexibility and responsiveness to the varied needs of the AI community in India.

To prevent potential misuse of subsidised compute resources, such as the trading of vouchers, the government plans to implement checks and balances.

Singh stressed the importance of ensuring that the subsidies are used appropriately and reach those genuinely in need of computational power for AI development. This safeguard is crucial to maintaining the integrity and effectiveness of the subsidy program.

Earlier, MeitY Secretary S Krishnan mentioned that to expedite and broaden the availability of GPU resources across the country, the government is considering mechanisms like viability gap funding or a voucher-based system.

The Government of India’s substantial investment in GPUs under the India AI Mission is a strategic move to bolster the nation’s AI capabilities. By making powerful computing resources more accessible and affordable, this initiative aims to foster innovation, support startups, and drive significant advancements in the AI sector.

The emphasis on user choice and prevention of misuse ensures that the benefits of this investment reach the right hands, ultimately contributing to India’s growth as a global AI powerhouse.