India's GPU Capacity Projected to Triple to 100,000 by 2026-End
India's installed graphics processing unit (GPU) capacity is on track to triple from the current 38,000 devices to approximately 100,000 by the end of 2026, according to a senior government official. This expansion is a key component of the government's strategy to develop a sovereign compute infrastructure and position India as a global leader in artificial intelligence (AI).
Government Initiatives Under the IndiaAI Mission
Abhishek Singh, Additional Secretary at the Ministry of Electronics and IT (MeitY) and CEO of the IndiaAI Mission, revealed that fresh tenders have been opened to significantly boost GPU installations. "By the end of the year, we should have roughly a 100,000 GPUs," Singh stated during an interaction with The Indian Express. This initiative is part of the Rs 10,370 crore IndiaAI Mission, which aims to subsidize GPU usage for Indian startups and researchers, making computing more affordable at around Rs 65 per hour instead of prohibitively high costs.
Computing is a critical layer for modern AI applications, enabling the training of massive neural networks, running complex algorithms, and processing real-time predictions. The government is not funding GPU procurement directly but offering discounts on infrastructure built by private players. Additionally, the government will become a user of GPUs for its own activities, owning the underlying hardware to augment sovereign capacity.
Sovereign Compute and Global Ambitions
Singh emphasized the need for sovereign GPU capacity, particularly for government work where AI loads must remain within government-owned infrastructure. "We are trying to augment a sovereign capacity for GPUs. We are augmenting the GPU capability of National Informatics Centre's (NIC's) Meghraj, and we have recently floated a tender for that," he explained. This move aligns with India's broader goal to become the "inferencing capital for the world," attracting AI workloads to gain a strategic advantage as the AI capital globally.
The recent Budget announcement of a tax holiday for foreign companies setting up data centers in India further supports this vision. Singh noted that the rationale is to position India as the data center capital of the world, leveraging AI to solve societal problems. "Our pitch is: build AI in India, for India, and once these solutions have scaled up, we will give an example to the whole world of how AI can be used for solving societal problems," he added.
Upcoming India-AI Impact Summit
Singh, who is spearheading the upcoming India-AI Impact Summit, highlighted that the event will address "every dimension of AI." This includes topics across the AI stack such as regulation, child safety, and more. "Compared to the previous three summits, this is much more comprehensive, bigger in scale, and will also put up the agenda of the Global South at the forefront," he said. The summit aims to foster discussions on AI-led solutions for on-ground problems, reinforcing India's commitment to ethical and impactful AI development.
In summary, India's GPU expansion under the IndiaAI Mission represents a significant step toward enhancing AI infrastructure, promoting innovation, and establishing the country as a key player in the global AI landscape. With a focus on sovereignty, affordability, and societal impact, these efforts are poised to drive transformative changes in technology and policy.
