Huang Renxun: Nvidia Compute Is "Interchangeable, Durable, High-Yield," Three-Year-Old H100 Chip Rental Prices Still Rising

On Monday, Nvidia CEO Jensen Huang reiterated his view that AI computing hardware is becoming a valuable, financeable asset as rental prices for the company's older H100 GPUs continue to climb.

Huang Renxun Calls Nvidia Compute an "Income-Generating Asset"

"Nvidia compute is interchangeable, durable, and offers extremely high rental returns," Huang wrote in a post on X. "It is a productive, income-generating asset."

Nvidia compute is interchangeable, durable, and offers extremely high rental returns. It is a productive, income-generating asset.

— Jensen Huang, September 8, 2026

His comments were in response to Ornn Exchange, which said its compute price index showed that rental prices for Nvidia's three-year-old H100 chip rose 22% in one month to $3.28 per hour.

The H100 is a three-year-old training chip. Its rental price rose 22% this month to $3.28 per hour.

Every depreciation model assumes a chip this old only loses value. The market is instead paying a higher price for it.

— Ornn, September 7, 2026

The increase is notable because traditional depreciation models typically assume older hardware loses value over time. However, strong demand for AI computing is helping some Nvidia GPUs maintain or even boost their rental economics.

Nvidia Wants Wall Street to View AI Chips as Infrastructure

Huang's comments build on Nvidia's broader push to turn AI compute into an investable asset class.

In August, Nvidia announced partnerships with Apollo Global Management, BlackRock Inc., Blackstone Inc., Brookfield Asset Management, Goldman Sachs, and KKR & Co. Inc. to help mobilize over $500 billion in funding for AI infrastructure.

The initiative aims to help AI labs, hyperscalers, and general enterprises tap institutional and private capital to finance data centers and Nvidia hardware.

Why Rising H100 Rental Prices Matter

These economics are critical to Nvidia's financing strategy.

If GPUs can consistently generate income through leasing, investors and lenders may increasingly assess them as productive infrastructure rather than traditional depreciating equipment.

However, FedWatch Advisors founder Ben Emons believes China is the biggest threat to Huang's push.

In August, Emons warned that China's rapidly growing domestic AI computing capacity could eventually flood the market with cheaper chips, triggering a price war that would undermine the value of Nvidia GPUs used as collateral.

Last year, legendary short-seller Jim Chanos warned that a key accounting issue involving Nvidia chips could pose "significant financial risk" to companies aggressively investing in the space.

Price action: Nvidia shares closed Friday at $230.36, up 0.84%, but fell 0.38% to $229.49 in after-hours trading. In Edge Rankings, Nvidia ranks in the 99th percentile for growth metrics while maintaining positive short-, medium-, and long-term price trend ratings. Available stock screeners allow comparison of its performance against industry peers such as AMD.