The Ministry of Electronics and IT (Meity) is set to seek cabinet approval for the India AI program, an initiative involving investments exceeding Rs 10,000 crore. Minister of Electronics and IT Rajeev Chandrasekhar disclosed that this program encompasses the establishment of GPU-based servers through a public-private partnership. The initiative is designed to address the increasing demand for GPU-based servers, renowned for their high-speed data processing capabilities compared to CPU-based servers.
Chandrasekhar, speaking at the inauguration of semiconductor-focused software company Synopsys, emphasized the program’s reliance on public-private collaboration. It will encompass computer infrastructure capacity development in both private data centers and public CDAC-run data centers.
India’s AI program seeks to bridge the gap in computing infrastructure for AI technology development, currently dominated by the United States and China. India’s strategy involves creating world-class AI computing infrastructure at five locations, offering 3,000 AI Petaflops computing power. This capacity represents a significant leap beyond current capabilities.
In addition to this, recommendations include establishing an Inference Farm and Edge Compute systems. NVIDIA, with an 88% market share, currently dominates the GPU market. Due to its global demand, there is a substantial lead time of 12-18 months for GPU procurement from NVIDIA.
The government has already invested Rs 1,218.14 crore in setting up 24 PetaFlops compute capacity under the National Supercomputing Mission over the past eight years.
Chandrasekhar highlighted the critical role played by companies like Synopsys in driving innovation within India’s technology and digital ecosystem. Synopsys has established a significant design center in India, housing around 6,000 engineers and contributing significantly to global design efforts.