Nvidia’s $500 Billion AI Plan Hinges on One Risky Bet: Chip Depreciation

Nvidia CEO Jensen Huang is proposing a bold financial strategy: using the company’s graphics processing units (GPUs) as collateral to secure $500 billion in new funding. The core idea is that these chips—the engines of the current artificial intelligence boom—hold enough long-term value to back massive loans. However, the entire plan rests on a single, fragile assumption: that the chips will not lose their worth too quickly. GPUs are not like real estate or gold. They are high-tech products that face rapid obsolescence. As newer, more powerful models are released, older chips lose value at a steep pace. If the AI market cools or a competitor produces a superior product, the collateral behind the loan could shrink dramatically. This creates a clear risk for lenders, who must decide whether the hardware’s future value justifies the enormous sum. Huang’s pitch is straightforward: lend against the machines that power the AI revolution. The problem is timing. The speed of technological change is the biggest variable. If depreciation is faster than expected, the financing plan could collapse under its own weight. For now, the market is watching to see if investors accept the gamble—or demand a different kind of guarantee.