India Steps Up AI Push with 100,000 Public GPUs Target by December

NEW DELHI, Sept 28: India is accelerating efforts to expand its artificial intelligence computing infrastructure, with the government targeting 100,000 public GPUs by December as part of its broader push to build domestic AI capabilities.

The expansion is aimed at increasing access to high-performance computing resources for researchers, startups and institutions developing AI applications. The initiative is also intended to reduce dependence on overseas cloud infrastructure at a time when access to advanced computing has become a central requirement for developing and deploying sophisticated AI systems.

According to a report published on September 28, the country’s AI Mission is stepping up the deployment of subsidised public GPUs, with the December target representing a significant expansion of computing resources available through the national programme. The initiative also includes a fellowship programme for 13,500 scholars, linking computing infrastructure with the development of specialised talent.

The availability of computing power has emerged as one of the most important factors shaping the global AI race. Training large AI models requires substantial processing capacity, while running such systems for millions of users also places considerable demands on data centres and specialised hardware.

For India, expanding public access to GPUs could help institutions that may otherwise find the cost of advanced computing prohibitive. Shared infrastructure can allow startups and research groups to experiment with AI models, develop sector-specific applications and test new technologies without having to independently build expensive computing facilities.

The move comes as India works to develop an AI ecosystem that extends beyond the use of foreign-developed models and platforms. Building domestic capacity involves several layers, including computing infrastructure, semiconductor technology, datasets, skilled professionals, research institutions and secure digital networks.

The government has increasingly highlighted the importance of these components as India seeks to expand its role in emerging technologies. Finance Minister Nirmala Sitharaman said on September 27 that greater investment was needed in AI infrastructure, semiconductor chips and quantum technology. She said the country needed stronger hardware capabilities, training and institutional capacity to take advantage of these technologies.

Sitharaman also pointed to the growing role of AI in India’s manufacturing and small-business sectors. She said enterprises, including MSMEs, would increasingly require AI-based solutions to improve productivity and remain competitive internationally.

The emphasis on infrastructure reflects a shift in the AI conversation. While much of the public attention has focused on chatbots and generative AI applications, the underlying computing infrastructure is becoming equally important. Data centres, processors, networking equipment, electricity supply and cooling systems are all necessary to operate AI services at scale.

A separate assessment published on September 28 highlighted the growing investment around India’s AI infrastructure. Companies involved in supplying power, data centres, equipment and other components needed for AI computing are increasingly becoming part of the country’s technology growth story.

The expansion of computing capacity could also influence the development of Indian-language AI systems. With access to more computing resources, researchers and companies can work on models designed for India’s diverse linguistic environment and applications in areas such as agriculture, education, healthcare, public administration and financial services.

The fellowship programme for 13,500 scholars is another important element because infrastructure alone cannot create a sustainable AI ecosystem. Advanced computing needs researchers, engineers and developers who can design models, optimise algorithms and create practical applications.

India’s technology strategy is therefore increasingly focused on combining physical infrastructure with human capital. Universities, research institutions, startups and established technology companies are expected to play different roles in developing the ecosystem.

The semiconductor industry is another closely connected area. AI systems depend heavily on advanced processors and memory technologies, making chip supply an important component of national technology planning. India has been seeking to expand domestic semiconductor manufacturing and develop capabilities across chip design, fabrication, packaging and related technologies.

The country’s semiconductor ambitions also face supply-chain and cybersecurity challenges. A recent assessment noted that India’s expanding chip ecosystem will need protection against cyberattacks, geopolitical disruptions and excessive dependence on individual suppliers. It called for greater diversification and stronger domestic capabilities in critical parts of the semiconductor chain.

As India increases public computing capacity, questions around energy consumption and data-centre infrastructure will also become more important. AI computing requires significant electricity and specialised cooling systems, meaning that the growth of digital infrastructure is closely linked to energy planning.

The World Economic Forum’s September 2026 Chief Economists’ Outlook said AI adoption is expected to accelerate over the next year, with economists anticipating productivity gains while also raising questions about whether investment in data centres will create employment on a comparable scale.

For India, the immediate objective is to ensure that access to computing does not become a barrier to innovation. The public GPU expansion is intended to make advanced resources more accessible while supporting the country’s wider effort to build domestic expertise.

The December target will therefore be an important milestone for India’s AI infrastructure programme. Its longer-term impact will depend not only on the number of GPUs deployed but also on how effectively researchers, startups, universities and businesses are able to use the computing capacity to develop useful and secure AI applications.