India’s AI Power Play: Infrastructure Opportunity

A new report from Shriram Mutual Fund, titled “The AI Bubble Debate: A Unit-Economics Lens,” posits a compelling, if unconventional, strategic pathway for India in the global artificial intelligence revolution. The report argues that India’s most significant opportunity lies not in developing advanced AI models, but rather in supplying the foundational electricity and infrastructure essential to power AI data centers worldwide.

This perspective frames the current AI investment cycle as fundamentally a “power story.” The core mechanism is simple: AI, at its root, is a massive computational endeavor, and computation demands colossal amounts of energy. A single one-gigawatt AI data center, for instance, can cost between $20 billion and $50 billion and consume electricity equivalent to approximately 750,000 homes. These facilities operate continuously, requiring not only vast power generation but also extensive transmission networks and energy-intensive cooling systems, making electricity the primary constraint on AI expansion, even more so than powerful GPUs or advanced chips.

The scale of capital expenditure underscores this structural reality. Big Tech’s AI capital expenditure from 2021 to 2025 reached an astounding $1.08 trillion, with an additional $725 billion projected for 2026. Looking further ahead, Goldman Sachs forecasts an even more staggering $5.3 trillion in AI-related infrastructure investment from 2025 to 2030. This monumental spend is directed overwhelmingly towards the physical infrastructure that underpins AI, rather than purely software development.

Despite India’s limited direct involvement in the frontier of global AI model development, the report highlights the country’s profound structural advantage. India can leverage its expanding electrical equipment manufacturing base and robust power infrastructure to become a critical supplier. This includes manufacturing and supplying essential components such as transformers, switchgear, cables, and grid infrastructure – the fundamental “picks and shovels” indispensable for every AI data center globally. This strategy offers stable investment opportunities irrespective of which specific AI platform or model ultimately achieves market dominance.

Addressing concerns about a potential “AI bubble,” Shriram Mutual Fund argues that current AI investments differ markedly from past tech bubbles. These investments are largely self-funded by highly profitable technology companies, shifting the debate from questions of solvency to those concerning the return on capital. This reframing suggests a more durable, if intensely capital-intensive, investment landscape.

Shriram Mutual Fund, reflecting this analytical stance, maintains an overweight position in the power sector, driven by the anticipated long-term demand stemming directly from AI infrastructure. The durable takeaway for India is clear: by focusing on its established industrial strengths in power and electrical infrastructure, the nation can strategically position itself as an indispensable enabler of the global AI revolution, moving beyond direct competition in model development to capture value from the underlying structural demand.