Nvidia’s rapid expansion has made a $10 trillion market value a conceivable target, but the companies driving much of the AI infrastructure boom are also developing alternatives to its hardware. Microsoft, Amazon, Alphabet and Meta are investing heavily in data centers while creating chips designed to reduce the cost of running selected AI workloads.
Data Center revenue rose 117% to $89 billion, and gross margin reached 75%.
The company’s latest results show that demand remains exceptionally strong, but they also raise the stakes for future competition. However, its largest customers have a direct financial incentive to use custom silicon when it can deliver comparable performance at a lower cost.
That shift is already visible across the cloud industry. Google has expanded its TPU program and introduced eighth-generation TPU systems for training and inference, while continuing to offer Nvidia-based infrastructure.
The financial question is whether Nvidia can preserve both its growth rate and its pricing power as custom chips handle a larger share of workloads. A $10 trillion valuation supported by a 25-times earnings multiple would require roughly $400 billion in annual net profit. Nvidia reported $120 billion in net income for fiscal 2026, meaning the valuation would depend on substantial additional earnings growth.
Cloud companies may continue using Nvidia systems for demanding workloads while deploying their own accelerators for specialized inference or other tasks. The key issue for investors is how much of the expanding AI-computing market Nvidia captures and how much profit remains with the chip supplier rather than its customers.