Google Aims to Double Computing Power Every Six Months to Meet Surging AI Demand

Google's cloud division faces an unprecedented infrastructure challenge as artificial intelligence adoption accelerates. The company now requires computing capacity to double every six months—a pace that outstrips even Moore's Law—to keep pace with industry demands. This revelation came during a November 6 all-hands meeting where Google Cloud VP Amin Vahdat outlined plans for 1,000-fold expansion within five years.

The tech giant is pursuing efficiency breakthroughs rather than blank-check spending to achieve this goal. Its seventh-generation Ironwood chip demonstrates the strategy, delivering 30x greater energy efficiency than 2018's inaugural model. Vahdat emphasized the critical constraint: this exponential growth must occur without proportional increases in energy consumption.

CEO Sundar Pichai separately warned employees of mounting pressures through 2026 as Google races to scale both AI and cloud infrastructure. The comments underscore the immense resource requirements fueling today's AI arms race, with implications for everything from data center construction to energy markets.