AI data centres are experiencing unprecedented power demand volatility that is straining equipment and threatening grid stability. According to Business Standard, AI data centres can see power usage spike as much as 50% above design capacity, with a 1 gigawatt facility potentially using 1.5 gigawatts for a split second. As Amber Villegas-Williamson from the Uptime Institute explains, "AI does create very unusual power demand, it's like over-revving your car wears out the engine faster than keeping a constant speed." The technical challenges are evident globally, with cracks developing in gas-fired turbines at facilities like xAI's Colossus computing facility in Memphis, Tennessee. Equipment failures are occurring at an alarming rate, with batteries sometimes needing replacement within months or even weeks due to the high strain from power fluctuations. As Bloomberg News reports, a gigawatt data center is equivalent to a city the size of Boston, half of which can flicker on and off every few seconds, creating repeated shocks that connected equipment struggles to absorb.
Solar power has demonstrated remarkable growth in India's energy landscape, contributing 14% of total power output in H1 2026, a significant increase from 9% in the previous year. According to reports from Upstox, this sharp growth is attributed to strong power generation capacity expansion and favourable government and state subsidies. The renewable energy sector is experiencing strong demand traction for solar power equipment, with companies like Premier Energies and Vikram Solar leading the market amid key earnings releases.
The artificial intelligence ecosystem is driving strong demand for data centres globally, with industry reports from KPMG India projecting solar power demand from data centres could reach 24 GW by 2030. As reported by Upstox, the current installed capacity stands at 1.9 GW with 4.5 GW in pipeline, requiring a capacity addition growth of 28.5% CAGR to meet projected demand. Data centres are mandated to fulfil minimum 51% energy requirements from renewable sources, creating institutionalised demand for solar power. This aligns with global trends where battery energy storage systems are increasingly deployed for data centre power support, representing a 3% share of the overall battery storage market.
The power grid faces significant challenges from AI data centre operations, with the North American Electric Reliability Corp (NERC) repeatedly warning that data centres are among the greatest risks to grid stability. According to Business Standard, NERC evaluated more than 33 gigawatts of operational data centres in the US and found about three quarters of their load models are insufficient to represent data-centre dynamic behaviour. The agency issued a rare level-three alert requiring big data centres to address immediate risks, with responses due by August 3. Power quality expert Sreemant Roy from Schneider Electric warns that "these loads are extremely dynamic or fluctuating, which causes grid instability and can lead to, if not corrected, potential blackouts or power outages." The problems are also a potential source of wider instability in power grids that are already straining to keep the lights on.
Data centres are emerging as significant customers for renewable power, creating multiplier demand for solar power generation. As reported by Upstox, a 1 GW of uninterrupted supply of power throughout the day would require a capacity of 4-6 GW, highlighting the substantial scale of power requirements. This demand creates long-term power demand visibility, with data centres signing long-term contracts for 10-20 years, providing stable revenue streams for solar power equipment manufacturers. The global battery energy storage market, valued at $52.40 billion in 2025, is projected to reach $187.90 billion by 2035 at a CAGR of 13.2%, indicating strong growth potential for energy storage solutions supporting data centre operations.
