Beyond compute: Power and cooling for AI data centers

McKinsey & Company projects AI-driven demand for data centers and the related capital and design implications.

• Main announcement/action: McKinsey & Company reports that AI-driven demand will expand data center capacity to 220 gigawatts by 2030 (a CAGR of 22%) and estimates $6.7 trillion in cumulative capital outlays worldwide by 2030 to meet compute demand; the report focuses on advances in power and cooling equipment and the need for codesigned systems.
• Background and details: The article emphasizes that power, cooling, and IT components must be codesigned, highlights opportunities for equipment manufacturers to pursue vertical integration and provide end-to-end services (repair, maintenance, commissioning), and notes time to market as a key buying preference for data center operators.