Nvidia Says Memory Pricing Has Turned ‘Extreme’ and Is ‘Headed Even Higher Into Next Year,’ But Its Own Price Increases Are Already Executed

Nvidia CFO Colette Kress warned investors that extreme memory pricing conditions have surpassed expectations and will rise further into next year, forcing the company to reset margin outlooks. High-bandwidth and server memory supply constraints, driven directly by the massive AI infrastructure buildout, are increasing input costs across the supply chain. In response, Nvidia has executed price increases for its server hardware to partially offset rising component expenses. The chipmaker reported a 75% gross margin for its latest quarter but guided margins lower to 74% in the subsequent quarter, projecting margins to bottom out between 71% and 72% in fiscal Q4 before stabilizing around 72% to 73% in fiscal 2028. Nvidia disclosed approximately $279 billion in supplier purchase commitments, with industry reports estimating that roughly $160 billion relates directly to memory suppliers Samsung, SK Hynix, and Micron. While CEO Jensen Huang affirmed sustained AI profitability and demand, margin compression highlights supply chain bottlenecks as fab expansion requires years. The margin reset reflects the trade-off of the AI boom, balancing premium hardware pricing power against severe memory scarcity.