NVIDIA reported record third-quarter revenue of $57 billion for the period ending October 26, 2025, reflecting a 22 percent sequential increase and a 62 percent jump from last year. The performance reflects NVIDIA’s dominant position in global AI infrastructure, data center innovation and next-generation computing platforms.
NVIDIA is targeting $65 billion in revenue for the fourth quarter of fiscal 2026. “Blackwell sales are off the charts, and cloud GPUs are sold out,” said Jensen Huang, founder and CEO of NVIDIA.
NVIDIA’s data center business remained the primary growth engine, generating a record $51.2 billion, up 25 percent from the previous quarter and 66 percent year-over-year. The company highlighted strong momentum for Blackwell architecture, which achieved top efficiency and performance scores in the SemiAnalysis InferenceMAX benchmarks. NVIDIA emphasized that Blackwell delivers ten times more throughput per megawatt compared with the previous generation, positioning it for large-scale AI factories and national AI infrastructure programs.
Major cloud and AI partnerships defined the quarter. NVIDIA and OpenAI agreed to deploy at least 10 gigawatts of NVIDIA systems to support OpenAI’s next-gen AI infrastructure. Leading technology companies, including Google Cloud, Microsoft, Oracle and xAI, committed to expanding U.S. AI capacity using hundreds of thousands of NVIDIA GPUs. Anthropic will run on NVIDIA infrastructure for the first time, adopting one gigawatt of compute powered by Grace Blackwell and Vera Rubin systems. NVIDIA also announced collaborations with Intel to co-develop custom data center and PC products utilizing NVLink.
NVIDIA is accelerating global supercomputing projects, working with Oracle to build Solstice and Equinox, two U.S. Department of Energy systems featuring up to 100,000 Blackwell GPUs. The company produced its first U.S.-made Blackwell wafer at TSMC’s Arizona fab, marking a milestone in reshoring semiconductor manufacturing.
The quarter also brought new AI platform launches, including Rubin CPX for massive-context processing, NVQLink for quantum-class GPU integration and expansions of Arm’s Neoverse platform with NVLink Fusion. Meta, Microsoft and Oracle will enhance their AI data center networks with Spectrum-X Ethernet switches, while BlueField-4 gained traction across leading cloud and cybersecurity providers.
NVIDIA announced multiple RAN and 6G-focused partnerships, including AI-RAN solutions with Nokia and an all-American AI-RAN stack developed with T-Mobile, Cisco, Booz Allen and others. The company also deepened collaboration with Palantir on operational AI systems and continued building next-generation AI infrastructure with partners in the U.K., Germany and South Korea.
In Gaming and AI PCs, NVIDIA posted $4.3 billion in quarterly revenue, largely stable sequentially but up 30 percent year-over-year. New game launches supported by DLSS 4 and Reflex, advances in RTX Remix and updates to 3D object generation tools reinforced NVIDIA’s leadership in AI-powered PC workloads.
Professional Visualization revenue reached $760 million, supported by the rollout of DGX Spark, a compact AI supercomputer that delivers the full NVIDIA AI software stack in a small form factor.
NVIDIA’s Automotive revenue grew to $592 million as the company introduced DRIVE AGX Hyperion 10 for autonomous vehicle development and partnered with Uber to target a fleet of 100,000 level-4-ready vehicles from 2027. The company also showcased progress in physical AI, collaborating with major industrial and robotics leaders to modernize U.S. manufacturing. New digital twin services from PTC and Siemens, along with the launch of IGX Thor for real-time edge AI, strengthened NVIDIA’s ecosystem in robotics and industrial automation.
Rajani Baburajan
