UBS currently expects AI capital expenditure to be close to 1 trillion US dollars in 2026 and further rise to about 1.4 trillion US dollars in 2027, and the main reason behind this increase is the sharp rise in memory costs. UBS estimates that memory-related spending will soar from $71 billion in 2025 to $367 billion this year, reaching $923 billion in 2027. The remaining AI-related costs are expected to be $631 billion in 2026 and fall back to $525 billion in 2027. This means that the increase in memory costs this year contributed about 60% of the increase in AI capital expenditure; by 2027, as spending on other components falls, the increase in memory costs will even exceed the total net increase in AI capital expenditure.

UBS currently expects AI capital expenditure to be close to 1 trillion US dollars in 2026 and further rise to about 1.4 trillion US dollars in 2027, and the main reason behind this increase is the sharp rise in memory costs. UBS estimates that memory-related spending will soar from $71 billion in 2025 to $367 billion this year, reaching $923 billion in 2027. The remaining AI-related costs are expected to be $631 billion in 2026 and fall back to $525 billion in 2027. This means that the increase in memory costs this year contributed about 60% of the increase in AI capital expenditure; by 2027, as spending on other components falls, the increase in memory costs will even exceed the total net increase in AI capital expenditure.