India’s administration has sanctioned a ₹10,300 Crore ($1.24 billion) funding package to strengthen the country’s artificial intelligence (AI) infrastructure.
At the core of this initiative is the proposal for a cutting-edge supercomputer equipped with a minimum of 10,000 GPUs. Although specific details about the supercomputer under the “IndiaAI Compute Capacity” umbrella remain undisclosed, the government anticipates the necessity of a public-private partnership for its construction.
In addition, a new educational institution, the “IndiaAI Innovation Centre,” is set to be established to spearhead the advancement and implementation of foundational models. This center will place particular emphasis on developing indigenous Large Multimodal Models (LMMs) and domain-specific models, with a focus on leveraging edge and distributed computing for optimal efficiency.
Furthermore, the funding will be allocated to three other key initiatives:
- The IndiaAI Startup Financing mechanism, aimed at facilitating funding to expedite the commercialization of startups and AI projects within the industry.
- The IndiaAI Datasets Platform, which will receive additional funding to enhance public sector datasets, enabling local AI enterprises and the government to access the necessary data for developing suitable AI applications.
- The IndiaAI FutureSkills program, designed to enhance the accessibility of graduate and postgraduate AI programs, and establish Data and AI Labs that offer fundamental AI courses across India, particularly in regions beyond major urban centers.
The primary objectives of this funding package are to promote technological self-sufficiency and ensure the democratization of AI benefits across all segments of society.
While it remains uncertain whether the planned supercomputer will achieve these objectives through the utilization of homegrown technology, India’s ambition to develop server-grade CPUs based on the RISC-V architecture is noteworthy. However, there is currently no evidence of such devices being produced in the country. Additionally, India lags behind in GPU development.
On the other hand, the focus on indigenous LLMs is promising, considering India’s linguistic diversity with 22 scheduled languages mandated to be nurtured by law. Notably, some languages, such as Bengali, Marathi, and Telugu, boast over 80 million speakers, whereas others have significantly fewer speakers. The emphasis on LLM development is crucial, especially for languages like Malayalam or Punjabi, spoken by approximately 35 million people, which may not be prioritized by major AI players.
India’s commitment to undertaking such endeavors independently is evident. However, the announcement lacks clarity regarding the specific private partners sought to expedite local AI advancements. India’s complex relationship with Big Tech, characterized by a blend of appreciation for local investments and stringent regulations, underscores the government’s efforts to create public digital assets aimed at preventing tech monopolies.
