SemiAnalysis projects that cumulative AI IT and datacenter capital expenditure from 2024 to 2029 will reach approximately $11.1 trillion (Save this).
Annual capex is expected to exceed $2 trillion by 2028, growing almost every single year in the forecast window with no sign of deceleration..
SemiAnalysis projects outstanding AI-related debt will reach approximately $7.1 trillion by 2029, second only to the US mortgage market.
That is a structured AI debt financing, the mechanism by which neoclouds, datacenter builders and infrastructure companies fund GPU deployments when they do not have Microsoft or Google's cash reserves.
Think of it like the mortgage market, but for compute.
Instead of borrowing against a house, companies are borrowing against contracted GPU capacity and datacenter lease agreements and lenders underwrite the cash flows from AI compute contracts and extend credit against them.
This has enormous systemic implications.
A $7 trillion AI debt market creates a new asset class, a new interest rate sensitivity, a new default risk category, and a new financial system dependency on AI revenue actually materializing.
If AI monetization stalls, the downside is not just lower stock prices but rather a credit event but in the meantime here is how you can benefit from all of this.
Micron (MU), SK Hynix, and Samsung are the only companies that can supply HBM at scale. The shortage is structural, pricing power is intact, and demand commitments extend years into the future. Micron specifically has triple-digit HBM revenue growth potential through 2026 and is the fastest-gaining share in HBM qualifications with Nvidia.