Hyperscalers commit nearly $2 trillion to AI hardware and memory capex; Google leads $811B surge, eclipsing consumer electronics spend.

Apple used to be among a handful of companies willing to purchase memory and other components worth billions of dollars under long-term supply contracts at fixed prices. The artificial intelligence era represents a new reality, marking a tectonic shift in the high-tech world. Alphabet, Microsoft, Meta, and Amazon now hold purchase commitments totaling approximately $2 trillion, with a significant portion dedicated to memory, according to analyst Claus Aasholm's estimates.

Combined purchasing commitments from these four major hyperscalers reached nearly $2 trillion by Q2 2026, with Alphabet and Microsoft accounting for the overwhelming majority. While these commitments are approximate, span many years, and should be taken with appropriate caution, they reflect the industry's fundamental direction.

This rapid expansion reveals several critical developments. The AI infrastructure race is accelerating rather than stabilizing. AI infrastructure investments are driven by a handful of hyperscale cloud service providers whose long-term procurement commitments now vastly exceed those of traditional consumer electronics companies such as Apple. Memory has become a strategic asset—perhaps a competition weapon—rather than a commodity. Suppliers of both 3D NAND and DRAM are gaining pricing power. Demand for memory will likely drive major capacity expansion at Micron, Samsung, and SK Hynix, even as these companies have remained exceptionally disciplined about capacity investments.

Google shows by far the most aggressive increase in purchasing commitments, rising from roughly $140–150 billion in Q3 2025 to around $811 billion by Q2 2026. Microsoft follows a similar trajectory, reaching approximately $678 billion in total obligations. Meta has ramped commitments substantially to around $349.3 billion, whereas Amazon increased more gradually to roughly $130 billion. By contrast, Apple—which makes the world's most popular smartphone and was the largest consumer of memory just a couple of years ago—remains nearly flat at approximately $57 billion, of which $56.2 billion is payable within 12 months. Apple's commitments fall well short of Nvidia's $119 billion.

While these figures represent total purchase commitments including foundry capacity, 3D NAND, and DRAM memory, a significant portion consists of contract manufacturing obligations. Alphabet, Amazon, Meta, and Microsoft all build custom silicon and custom servers, so substantial commitments go to various EMS providers. The chart encompasses memory and broader capacity at TSMC, Samsung Foundry, and GlobalFoundries—which have become strategic assets rather than merely another component to procure at the best available price.

AI infrastructure requires enormous quantities of AI accelerators, DRAM (including HBM), and 3D NAND. The supply of high-end memory is constrained by fab capacity at major DRAM makers, while AI accelerator supply is constrained by both wafer capacity and foundry and packaging capacity. Hyperscaler CSPs therefore have an incentive to lock in supply years ahead, even if doing so requires exceptionally large purchasing commitments. This fundamentally changes the relationship between semiconductor suppliers and their customers. A company guaranteeing hundreds of billions in future purchases can effectively underwrite expansions of foundry, memory, and advanced packaging capacity in return for priority access to scarce products and future process technologies.

Apple's position in these graphs is particularly striking: its purchasing commitments barely move while those of Alphabet, Amazon, Meta, and Microsoft surge. If the trend continues, Apple may remain one of the world's largest semiconductor buyers in absolute terms, but the question is whether it will be among the key customers that foundries, memory makers, and OSATs prioritize in their future capacity expansions.

The most interesting takeaway is that the industry's center of gravity has shifted. During the smartphone era, foundries and memory suppliers competed aggressively for Apple's business because of its enormous purchasing power. Today, hyperscalers building AI infrastructure are making purchasing commitments that dwarf those of traditional consumer electronics companies—a strategic inflection point akin to what Andy Grove described in *Only the Paranoid Survive*.

Whether this tectonic shift results in prioritization of customers capable of enabling future capacity expansions through massive long-term agreements, or whether foundries and memory makers remain disciplined with capacity expansion to avoid losses during demand declines, remains an open question. What is certain is that the AI megatrend has transformed semiconductors—from foundries to advanced packaging, and from DDR5 to HBM4—into strategic assets that can no longer be treated as ordinary components procured on demand.