NVIDIA designs the H100 and Blackwell data-center GPUs used for artificial intelligence.
NVIDIA’s graphics processing units, or GPUs, can perform many calculations in parallel, making them useful for training and running modern AI models. The H100 is based on NVIDIA’s Hopper architecture, while Blackwell is a later platform introduced for demanding generative-AI and high-performance-computing workloads.
NVIDIA began in 1993 as a company focused on graphics hardware for computers. Its CUDA software platform, introduced in 2006, allowed developers to use NVIDIA GPUs for general-purpose computation, helping the company build an ecosystem beyond video games.
The company’s market capitalization rose sharply as cloud providers and technology firms invested in AI infrastructure. A common mistake is calling NVIDIA primarily a chip manufacturer: it designs chips and systems, while much of its semiconductor manufacturing is outsourced to foundries such as TSMC.