Google's computing power bill is rising at an alarming rate. Alphabet's second-quarter financial report shows that capital expenditures jumped 100% year-over-year, reaching a record $44.92 billion. At this pace, the annualized spending is already approaching $180 billion. But the real story is the ever-accelerating money-making machine behind this expenditure.

image.png

Google's revenue in the second quarter grew 24% year-over-year, reaching $119.8 billion, exceeding analysts' estimates of about $116.9 billion. The most impressive was Google Cloud — revenue surged 82% to $24.8 billion, with operating profit margin nearly doubling to 35.6%, and backlogs have now reached $514 billion. Gemini also delivered progress: monthly active users reached 950 million, and the model API processes as many as 22 billion tokens per minute. As cloud business handles massive demand while expanding profit margins, it is the direct confidence behind Google's decision to double its capital spending.

This wave of AI infrastructure isn't limited to Google alone. Morgan Stanley predicts that capital expenditures by hyper-scale cloud service providers will jump from around $800 billion in 2026 to $1.2 trillion in 2027; Goldman Sachs estimates $76.5 billion for 2026, rising to $1.6 trillion by 2031. McKinsey's projection is even more dramatic — the total capital expenditures related to this will reach $6.7 trillion by 2030, with data centers focused on AI alone absorbing $5.2 trillion. When a company can double its quarterly capex and simultaneously double its cloud profit margin, it's clearly not betting on returns from just one or two quarters, but rather on an entire infrastructure cycle rewritten by AI.