Discuss the trade-offs between air-cooling and liquid-cooling technologies for hyperscale data centres, with reference to India's growing digital infrastructure needs.

India's data centre capacity grew from about 375 MW in 2020 to nearly 1,500 MW by 2025 [2], and cooling — which absorbs roughly 7% of power in efficient hyperscale sites but over 30% in inefficient ones [3] — has become the decisive design choice. A 1-GW facility must continuously reject close to 1 GW of heat [4], forcing a trade-off between water and energy.

Why cooling is the binding constraint

• Almost all electricity drawn by servers is converted into heat through resistive and switching losses across millions of processors [4].
• AI accelerators have raised rack power densities sharply, pushing conventional cooling towards its physical limits [4].

Case for air-cooling

• Minimal water withdrawal — critical in water-stressed districts; Google's proposed 1-GW Visakhapatnam facility opted for air-cooling after sustained local protests over water use [4].
• Lower capital cost and a mature, widely serviceable technology.
• Trade-off: it imposes an efficiency and performance "tax" — more fan energy, or throttled compute — rather than a cost tax [4].

Case for liquid cooling

• Liquid carries heat away far more effectively than air, enabling denser, higher-power racks and lower cooling energy per unit of compute [3].
• Closed-loop direct-to-chip and immersion systems avoid the heavy evaporative water draw of older water-cooled designs.
• Trade-off: higher upfront investment, retrofitting difficulty, and dependence on specialised maintenance skills.

Weighing it for India

• Projects like the Visakhapatnam hub — around 600 acres and an expected ₹10,000 crore for Andhra Pradesh [1] — deliver jobs and digital sovereignty, so cooling choices must not become a growth barrier.
• The rational path is site-specific: air-cooling in water-scarce regions, closed-loop liquid cooling where power efficiency dominates, with mandatory water- and energy-use disclosure.

Neither technology is universally superior; each converts one scarcity into another. As India scales AI-era infrastructure, cooling policy should be integrated with state water budgeting, renewable power procurement and transparent reporting, so that digital growth advances SDG-6 on water security alongside SDG-9 on resilient infrastructure.