Nvidia's AI Infrastructure Boom Mirrors Unsustainable 1960s Computer Leasing Practices

Renowned investor Michael Burry has sounded an alarm about the state of artificial intelligence infrastructure, warning that it mirrors unsustainable computer-leasing practices from the late 1960s.

Burry pointed to a slide from Nvidia's September investor presentation that showed AI hardware retains high residual value beyond standard five-year depreciation schedules. However, he claims this data is based on a discounted cash flow model rather than actual resale prices for second-hand equipment.

Burry notes that the comparison between an eight-year cash flow model and accelerated depreciation curves creates an 'apples-to-oranges' illusion. He argues that Nvidia's reliance on accelerated depreciation curves implies how severely each new product generation devalues its predecessors.

The investor also highlights that recent rebounds in older GPU rental rates stem from global memory and energy bottlenecks, not the hardware's intrinsic durability. Burry predicts that once these underlying component shortages ease, rental prices for legacy hardware will drop significantly.

Burry likens the current AI boom to the 1968 computer leasing crisis, where third-party companies offered extended depreciation schedules that ultimately led to industry-wide insolvencies when IBM launched its next-generation System/370. He warns that a similar dynamic is unfolding across modern neoclouds and private equity-backed data centers.

Burry has repositioned his portfolio, replacing short common-stock positions with put options, including September 2027 puts at strikes in the mid $100s for Nvidia stock.