Nvidia

Nvidia Corporation is an American accelerated computing company founded in April 1993 and headquartered in Santa Clara, California, that designs graphics processing units (GPUs), system-on-a-chip units, and the CUDA parallel computing platform. For its fiscal 2025 (ending January 26, 2025), Nvidia posted $130.5 billion in annual revenue, with data center sales of $115.2 billion, a 142% year-over-year surge driven by AI infrastructure demand. Its H100 “Hopper” GPU, launched in late 2022, and the Blackwell GPU architecture, unveiled at GTC on March 18, 2024, underpin frontier AI training; in the August 2024 MLPerf Training v4.0 round, Nvidia systems set records across all eight enterprise benchmarks. According to Jon Peddie Research, Nvidia held a 90% share of the discrete desktop GPU market in the third quarter of 2025. The company’s supply chain depends on advance agreements with SK hynix and Micron for HBM3e memory, securing capacity through 2026. Nvidia matters now because its proprietary CUDA ecosystem, rapid architecture cadence, and advance procurement agreements for high-bandwidth memory create a vertically integrated stack that competitors cannot replicate, making it the indispensable engine for frontier AI model training and inference across every major cloud provider and AI lab.

Two-hop subgraph: this entity, every entity it directly relates to, and every entity those neighbors relate to. Drag a node, scroll to zoom, click to inspect — or click any neighbor and re-center the atlas there.