GS-III: Economy, Environment, S&T
AI Compute Infrastructure and Data-Centre Capex Economics
No question on this topic yet. The note is still here to read.
Capital intensity, land availability, uninterrupted power supply, and water availability for cooling jointly determine where AI training infrastructure is built, making infrastructure economics the axis that separates sovereign compute ambitions from software or algorithmic questions. GPU cluster procurement, hyperscale and sovereign data-centre construction, PUE metrics for power and cooling, and fibre and low-latency interconnects are the core components. India's IndiaAI Mission and a Rs 10,000 crore AI corpus announced in Union Budget 2024-25 reflect the sovereign-compute dimension. MeitY's Data Centre Parks Scheme and India's National Data Centre Policy are the operative policy instruments framing this investment calculus.