Global data center capex surged past $170 billion in a single quarter as hyperscalers accelerate AI buildout.

Amazon, Alphabet (Google's parent), Meta, and Microsoft collectively spent $170 billion on data centers in a single quarter, representing a 72% increase from the same period last year. Wall Street now estimates their combined capital expenditure at roughly $1.5 trillion through the end of 2027.

Amazon led the infrastructure spending surge in Q2 2026 with $53 billion in capex, a 69% year-over-year jump. The company raised its full-year guidance to approximately $220 billion, citing soaring memory chip costs as a key driver. Meta reported a 55% year-on-year increase in infrastructure costs and revised its 2026 capex target upward to $130 billion.

Component price inflation alone added tens of billions to hyperscaler budgets during the quarter. This acceleration follows a pattern established in Q1 2025, when industry capex already reached $134 billion, up 53% year-over-year.

All four companies have pledged investments in new power generation capacity to sustain their expanded operations. This infrastructure buildout extends beyond data centers—Nvidia, AMD, and other semiconductor firms are allocating massive portions of their output to hyperscaler contracts, a supply chain that also feeds cryptocurrency mining hardware and blockchain infrastructure.

Bitcoin miners, particularly publicly traded operations like Marathon Digital and Riot Platforms, face a tightening hardware market as a result. Some mining companies, including Core Scientific and Hut 8, have struck deals to repurpose facilities for AI workloads, blurring the lines between mining and data center operations.

At a combined annual run rate of $680 billion, the four hyperscalers' capital deployment must flow from operating cash flow or debt markets, where it competes with every other borrower for available capital. Their investments in power generation and long-term energy contracts reshape the energy landscape for all market participants.

The investment thesis hinges on execution: if these massive infrastructure outlays support sufficient revenue and margins over the next decade, they represent essential infrastructure for AI applications. Conversely, if revenue doesn't materialize proportionally, the industry faces the most expensive stranded assets in corporate history.

The relationship between AI data centers and Bitcoin mining facilities has become increasingly symbiotic, with shared power infrastructure and even overlapping physical locations, creating an interconnected ecosystem where Big Tech's infrastructure decisions directly affect energy availability for other industries.