AI infrastructure spending to hit $11.1 trillion by 2029, Wall Street

AI infrastructure investment is projected to total $11.1 trillion through 2029, with annual outlays topping $2 trillion by 2028—far higher than Wall Street’s recent pessimism suggests [1]. The scale of financing, including an estimated $7.1 trillion of AI‑related debt, signals a deepening asset class that could reshape semiconductor supply chains.

Why the forecast matters

SemiAnalysis’s outlook rests on hyperscalers locking in long‑term GPU and datacenter agreements, which turn predictable cash flows into collateral for lenders. This financing model creates a new “AI compute” asset class, distinct from traditional equity exposure, and could sustain spending even if revenue growth slows. The projected debt level would rank second only to the U.S. mortgage market, underscoring the breadth of credit flowing into AI infrastructure [1].

Supply‑chain ripple effects

Spending will cascade through the entire semiconductor ecosystem. Companies that produce GPUs, memory chips, and the equipment to fabricate them stand to benefit, as each dollar of AI spend passes through multiple layers before a model generates output. The report highlights that firms such as Nvidia, Taiwan Semiconductor, Micron, and AMD occupy critical positions in this expanding chain, suggesting a prolonged demand horizon for hardware suppliers [1].

Market sentiment vs. fundamentals

Wall Street analysts have flagged AI capex as a potential bubble, pointing to soaring memory prices and high‑cost projects that exceed current profits of major cloud players [3]. However, the forecast argues that the underlying contracts and the necessity of AI for competitive advantage keep the spending trajectory robust, challenging the view that AI growth is a late‑cycle, inflationary drag [2][3].

What to watch

• Debt issuance: Monitor the volume of AI‑linked bonds and loans as lenders expand credit against GPU contracts.
• Semiconductor earnings: Track quarterly results of key hardware suppliers for signs of demand acceleration or slowdown.
• Capex announcements: Watch hyperscaler disclosures of new AI‑focused datacenter projects, especially any deviations from the multiyear contract trend.

The divergence between Wall Street’s slowdown narrative and the projected multi‑trillion‑dollar AI spend raises a key question: will the financing mechanisms and long‑term contracts sustain the boom, or could a shift in demand expose the growing debt exposure? Only forthcoming earnings and credit data will clarify the trajectory.