Nvidia’s $500B Bet To Make AI Compute Wall Street’s Next Asset Class - Forbes

NVIDIA, alongside major financial institutions like BlackRock and Goldman Sachs, has announced plans to raise over $500 billion to fund AI 'factories'—data centers that train and run AI models—with th... Read original →

NVIDIA’s push with Wall Street to institutionalize AI compute marks a fundamental shift in how data centers are valued and financed. This move will intensify demand for GPUs like H100 and A100, spurring upstream investments in EUV lithography, silicon wafers, and packaging. However, the rapid pace of technological obsolescence creates a mismatch between financing terms and hardware lifecycle, especially with loan structures not accounting for fast-evolving AI compute needs. Amazon’s recent adjustment to server depreciation timelines signals industry-wide risk exposure. Without long-term contracts and stable cash flows, AI factories may face significant write-downs. In the next 12–24 months, success will hinge on whether financial models can align with the dynamic nature of AI hardware. The key differentiator will be the ability to lock in revenue streams through standardized agreements, avoiding premature capital losses.