Nvidia Denies Pausing Take-or-Pay AI Compute Partnerships

Nvidia denied reports that it paused transactions under its take-or-pay AI Compute Partnership initiative. The company responded to a Wall Street Journal report claiming that deals were put on hold due to internal antitrust concerns and backlash from cloud providers.

The dispute reportedly involved Nvidia attempting to control how partners leased GPU capacity, specifically by requiring that providers lease only to Nvidia-approved customers and prevent single large clients from dominating capacity. Nvidia maintains that the business model remains in place and continues to evolve to meet high demand.

Under the program, Nvidia provides demand commitments and minimum revenue guarantees to help partners secure financing for AI data centers. As of July 26, 2026, Nvidia had committed $36 billion through six-year agreements. This structure allows the company to generate revenue from initial hardware sales and a subsequent share of rental income.