NVIDIA Disrupts Tech Supply Chain With New GPU Revenue Sharing Model - HotHardware

NVIDIA is disrupting the tech supply chain with a new GPU revenue-sharing model that goes beyond traditional hardware sales. Unlike the conventional approach where companies buy processors, build data... Read original →

NVIDIA’s GPU revenue-sharing model fundamentally shifts AI infrastructure economics by converting capex into opex. Technically, this accelerates adoption of Grace Blackwell architectures among smaller cloud providers and forces deeper integration with NVIDIA’s software stack—undermining AMD and Intel’s compatibility plays in inference. Regulatory risks loom: tying hardware to ongoing revenue may trigger antitrust scrutiny in the U.S. and EU, especially as performance-linked data flows cross borders under tightening export controls. Competitively, AMD lacks the CUDA moat to replicate this model; TSMC (Taiwan, China) benefits from GB300 demand but faces customer concentration exposure. Over the next 12–24 months, 'GPU-as-a-service' will proliferate, yet over-financialization could destabilize the ecosystem—if AI monetization lags, cash-strapped cloud startups may default, turning NVIDIA’s recurring revenue into a liability.