Nvidia’s first-quarter earnings for fiscal year 2026 revealed a company capitalizing on surging demand for artificial intelligence infrastructure, even as it faces mounting geopolitical challenges.
The chipmaker reported USD 44.1 billion in revenue for the quarter, up 69% year-on-year (YoY). The data center segment accounted for the bulk of that performance, growing 73% to USD 39.1 billion on the back of accelerating AI inference workloads and the rollout of its new Blackwell architecture.
The quarter marked the commercial debut of Blackwell systems, following their introduction earlier this year. These systems, including the GB200 NVL72, represent a significant leap from the prior Hopper generation, pairing Grace CPUs with Blackwell GPUs to deliver higher throughput and lower inference costs. “Grace Blackwell NVLink72 is the ideal engine today, the ideal computer thinking machine […] for reasoning AI,” said CEO Jensen Huang, citing up to 40 times the performance of its predecessor.
The broader shift toward AI inference is being felt across the tech sector. Microsoft, for example, reported a fivefold increase in tokens processed on its Azure OpenAI platform over the past year, according to Huang. The trend points to surging infrastructure demand, a backdrop that has translated into substantial hardware orders for Nvidia.
“On average, major hyperscalers are each deploying nearly 1,000 NVL72 racks or 72,000 Blackwell GPUs per week and are on track to further ramp output this quarter,” said Colette Kress, Nvidia’s CFO. “Microsoft, for example, has already deployed tens of thousands of Blackwell GPUs and is expected to ramp to hundreds of thousands of GB200s, with OpenAI as one of its key customers.”
However, Nvidia’s momentum was partially offset by new US export restrictions that cut off its data center GPU sales to China. The company recorded a USD 4.5 billion charge in Q1 to account for unsellable inventory and purchase obligations tied to its China-specific H20 chips. The hit dragged earnings per share (EPS) down to USD 0.76 on a GAAP basis and USD 0.81 on a non-GAAP basis. Excluding the charge, non-GAAP EPS would have been USD 0.96.
“The USD 50 billion China market is effectively closed to US industry,” Huang said during the earnings call. “Shielding Chinese chipmakers from US competition only strengthens them abroad and weakens America’s position. Export restrictions have spurred China’s innovation and scale. The AI race is not just about chips. It’s about which stack the world runs on.”
Looking ahead to Q2, Nvidia expects total revenue of USD 45 billion, with continued growth in data center sales outside of China. Gross margins are projected to reach 72%, buoyed by improving economics tied to Blackwell system deployments.
The company is also advancing a new category of infrastructure it refers to as “AI factories.” They refer to dedicated platforms for training and serving AI models at scale. These vertically integrated systems are designed for large-scale token generation, reasoning, and real-time deployment. Kress noted that approximately 100 such AI factories are currently in development.
Nvidia’s networking business plays a critical role in enabling this model. Revenue from networking rose 64% sequentially to USD 5 billion, with NVLink shipments topping USD 1 billion. Spectrum-X, the company’s enhanced Ethernet platform, continues to gain traction and is now annualizing at over USD 8 billion. New customers this quarter included Google Cloud and Meta.
Other segments also showed strength. Gaming revenue hit a record USD 3.8 billion, boosted by the launch of the GeForce RTX 5060 and 5060 Ti GPUs. Automotive revenue rose 72% YoY to USD 567 million, supported by new deployments with General Motors and Mercedes-Benz.
In industrial applications, Nvidia’s Omniverse and Isaac platforms are seeing broader adoption. Omniverse has been implemented by companies such as Kion Group, Pegatron, and Accenture, while Isaac has found users in Agility Robotics, Boston Dynamics, and Xpeng.
Despite regulatory headwinds, Nvidia’s results suggest it remains well-positioned to lead the next phase of AI infrastructure, one that’s defined not just by hardware capabilities, but by the scale and integration of the systems it enables.