|
With 70 per cent of compute-intensive models being developed in the United States, China's growing push for technological isolation, and Western markets focusing on regulatory stability, India is emerging as a credible alternative for long-term digital infrastructure partnerships, according to a report by Centrum.
The report highlighted that the scale of India's opportunity is already becoming visible. The country's data centre capacity stood at 1.4 GW in 2024, representing a market size of USD 5.03 billion. This capacity is projected to grow at a compound annual rate of 21 per cent through 2030, almost double the global average growth rate of 11.2 per cent. Currently, the pipeline includes 3.4 GW of data centres under construction with more than USD 50 billion in hyperscaler commitments. However, the report points out a significant demand-supply gap in the coming years. Demand projections suggest that India will require between 6.5 GW and 8.3 GW of capacity by 2028, while the projected supply stands at only 4.8 GW. This imbalance, the report noted, is likely to create premium pricing power for early investors, especially as artificial intelligence (AI) workloads are expected to drive a 165 per cent increase in data centre power demand by 2030. Industry validation of this growing potential comes from several major investments already announced. Amazon Web Services (AWS) has a USD 12.7 billion investment plan in India, Google is building a USD 6 billion facility in Visakhapatnam, Microsoft is expanding with a USD 3 billion plan, and Reliance along with NVIDIA is developing a USD 25 billion AI infrastructure project in Jamnagar. The report further highlighted that India's strong digital economy supports this growth. The country already processes 164 billion annual UPI transactions, has over 547 million OTT users, and records nearly 17.4 exabytes of monthly data consumption, reflecting the massive scale of digital engagement and cloud demand. The report stated that India now stands at the epicenter of a historic opportunity to capture disproportionate value from the global AI infrastructure revolution, which is set to reshape competitive dynamics for the next generation. Unlike earlier technology cycles that provided optional upgrades, the report emphasizes that artificial intelligence represents an "existential necessity." AI workloads have expanded massively - from GPT-1's 117 million parameters to today's frontier models requiring over 30 trillion tokens. This scaling has transformed infrastructure needs, shifting from traditional 4-8kW per rack to AI training demands of 30-120kW per rack. This has led to the emergence of distinct market segments such as high-density "AI factories" for training and distributed inference centres. (ANI)
|
