India's data-centers need $80B by 2030 to meet AI demand
What's the story
India's data center industry is gearing up for a major investment drive, driven by the rising demand for cloud and artificial intelligence (AI) computing. Amit Sarin, the Managing Director of Anant Raj, and Darshan Hiranandani, Chairman and Co-Founder of Yotta Data Services, both agree on this trend. Hiranandani noted that India currently has about 1.5GW of conventional cloud capacity, which needs to double by 2030, even before accounting for AI workloads.
Investment requirement
$40B needed to double India's cloud capacity
Hiranandani emphasized that to double India's conventional cloud capacity from 1.5GW to 3GW by 2030, an investment of around $40 billion is required. This investment would go toward building infrastructure such as power systems and servers. He also said if India were to reach a capacity of 5GW by the end of this decade, total capital expenditure could rise up to $70-80 billion.
Development timeline
Infrastructure development requires time
Hiranandani stressed that infrastructure development for data centers can't happen overnight. He said power systems, servers, as well as cloud systems take years to deploy. Despite this long timeline, he expects demand to remain strong through 2030 with minimal risk of overcapacity in India due to its low starting base.
Value chain
Real estate the 1st step in data center value chain
Sarin said real estate is just the first step in the data center value chain, with the long-term returns coming from operating models like co-location and cloud services. He revealed Anant Raj has 157MW of ready data center buildings and around 28MW of operational co-location capacity. Sarin expects this capacity to rise to 63MW next year and further up to 117MW after that.
Demand forecast
AI's affordability won't reduce data center demand
Both Sarin and Hiranandani agreed that if AI becomes cheaper, the data consumption will increase. They said even without global AI demand, India's domestic cloud needs alone can support several years of capacity expansion.