US Hyperscalers' AI Spend Strains as DRAM Demand Soars

US hyperscalers are heavily investing in AI infrastructure, leading to increased capital expenditure consuming nearly all of their operating cash flow. This trend benefits DRAM suppliers, who gain pricing power due to limited supply. However, concerns arise about potential overinvestment and challenges in achieving sustainable AI profitability.

As US hyperscalers embark on a massive spending spree on AI infrastructure, Jefferies reports signs of financial strain. Capital expenditure by these tech giants is now absorbing nearly all their operating cash flow, even as DRAM suppliers benefit from unprecedented pricing power.

Projections by Jefferies indicate that by 2026, US hyperscalers will spend 92% of their operating cash flow on capital expenditures, a significant jump from 41% in 2023. This reflects the fierce AI arms race, with total annual capex expected to reach $700 billion this year and $800 billion next year, equating to 2% of the US GDP.

The surge in spending is flowing towards memory chips, particularly DRAM, as hyperscalers potentially allocate 28% of their cash flow to it. With Moore's Law losing steam, the DRAM market now experiences supply constraints, prompting companies like Nvidia to secure long-term supply deals amid fears of overinvestment and questions surrounding AI monetization.