The Indian government is preparing to announce the winning bidder for the procurement of 10,000 Graphics Processing Units (GPUs) under the IndiaAI Mission. This announcement is expected by the end of the week, according to a senior government official. This procurement is a crucial step in strengthening India’s artificial intelligence (AI) ecosystem, enabling startups, research institutions, and enterprises so that they can access high-performance computing resources which are essential for training AI models.
Alongside this, the government is finalizing a dedicated portal that will allow startups and researchers to access these high-performance GPUs at affordable rates. “The government has been working on developing the portal, and we are almost done with it,” the official said. This initiative aims to offer cost-effective infrastructure for AI training, empowering businesses and institutions to advance their artificial intelligence (AI) projects more effectively.
The bidding process, which took place on January 22, saw participation from ten major firms, including Jio Platforms, Tata Communications, E2E Networks Limited, and Yotta Data Services Limited. The government’s objective is to ensure a competitive and efficient allocation of resources, optimizing the available infrastructure for AI development.
The IndiaAI Mission, backed by a budget of Rs 10,371.92 crore, has allocated approximately 44 per cent (Rs 4,563.36 crore) for expanding computational capacity over the next five years. By providing access to 10,000 GPUs, the government aims to address the critical shortage of AI computing power in India, a necessity for training large-scale models in natural language processing, computer vision, and deep learning applications.
This initiative is expected to particularly benefit AI startups and research institutions, which often struggle to afford high-end GPUs due to their exorbitant costs and limited availability. The centralized procurement model ensures that compute power is made available at subsidised rates, optimising usage across sectors such as healthcare, fintech, agriculture, and cybersecurity.
The procurement of GPUs comes at a time when recent AI advancements are reshaping the way AI models are built. A notable example is Chinese startup DeepSeek’s open-source reasoning model, R1, which has demonstrated that cutting-edge AI can be developed without large-scale GPU access. However, the government remains confident in its AI infrastructure strategy. “IndiaAI’s compute pillar is designed to be open and flexible. The goal is to ensure that institutions and researchers have access to GPUs and can drive innovation,” the official stated. This suggests that while AI development methods may evolve, the need for a robust AI infrastructure remains crucial.
The government has also noted that since DeepSeek’s model is open-source, Indian startups and researchers are free to leverage it for their own AI developments.
A New Approach to GPU Procurement
Unlike traditional funding models, where the government provides financial assistance for infrastructure development (viability gap funding), the Ministry of Electronics and Information Technology (MeitY) has opted for a bulk procurement model. This approach allows both established companies and new entrants to compete for bids, ensuring maximum utilization of available resources.
Additionally, the ministry is considering subsidies for GPU usage in specific areas such as research and innovation, ensuring that cutting-edge AI technology remains accessible to startups and academic institutions.
With the successful bidder set to be announced soon and the AI compute portal nearing completion, India is taking a significant step toward creating a self-sufficient AI ecosystem that fosters innovation and research.
