As the global race to build artificial intelligence infrastructure gathers pace, a new report by Shriram AMC suggests that the biggest investment opportunity may not lie in AI software or semiconductor companies, but in the power sector. The report argues that electricity—not advanced chips—is emerging as the most critical factor driving the future of AI.
Titled “The AI Bubble Debate: A Unit-Economics Lens,” the report highlights that while graphics processing units (GPUs) account for a significant share of AI data centre costs, the availability of reliable electricity has become the real bottleneck. According to the report, a single AI data centre with one gigawatt of capacity can cost between $20 billion and $50 billion while consuming as much electricity as nearly 750,000 households.
Shriram AMC points out that global technology giants have already committed enormous investments to AI infrastructure. The world’s leading hyperscalers are estimated to have invested around $1.08 trillion in AI-related capital expenditure between 2021 and 2025, while spending in 2026 alone is projected to reach nearly $725 billion. Goldman Sachs estimates total AI infrastructure investments could rise to $5.3 trillion by 2030.
Despite this massive spending, the report believes the industry’s greatest challenge is shifting from semiconductor availability to power generation and transmission. It argues that regardless of which AI company or model dominates the market, every AI system ultimately depends on a stable and scalable electricity supply.
For India, the report sees the opportunity in infrastructure rather than AI model development. Since the country has limited listed companies involved in building frontier AI models or hyperscale cloud platforms, investors may find greater value in businesses that support AI infrastructure.
Shriram AMC maintains a positive outlook on India’s power ecosystem, including power generation companies, transmission utilities, EPC firms, power financiers, transformer manufacturers, switchgear producers, cable manufacturers, diesel generator companies, and cooling infrastructure providers. These sectors are expected to benefit as investments in AI data centres continue to expand globally.
The report also suggests that investing in infrastructure could be a relatively safer strategy than betting on individual AI companies. It notes that while the long-term profitability of AI business models remains uncertain, the demand for electricity will continue to grow as AI adoption increases.
According to the report, the AI industry would require an estimated $600-650 billion in additional annual revenue to generate a 10% return on current investments, whereas today’s AI revenue is estimated at only $50-150 billion. This indicates that while AI monetisation may still take time, infrastructure development is likely to continue because it is backed by some of the world’s largest and financially strongest technology companies.
Shriram AMC concludes that India’s biggest opportunity in the AI revolution lies in supplying the infrastructure that powers artificial intelligence. Rather than searching for the next breakthrough AI platform, investors may benefit more from companies involved in electricity generation, transmission networks, transformers, cables, electrical equipment, and cooling systems—the essential backbone that keeps AI data centres running.
Recent Random Post:
