Nvidia designed the H100 graphics processor used in many generative-AI data centers.
The H100 is a data-center graphics processing unit, or GPU, based on Nvidia's Hopper architecture. It was developed for demanding workloads including artificial intelligence training, inference, scientific computing, and high-performance computing.
Nvidia does not generally manufacture its most advanced chips in its own fabrication plants. Instead, it designs processors and relies on foundry partners, especially Taiwan Semiconductor Manufacturing Company, for production. This distinction explains why Nvidia and TSMC can both appear among the world's largest companies by market capitalization while playing different roles in the semiconductor supply chain.
Demand for accelerated computing helped Nvidia's market value rise sharply in the 2020s. Its CUDA software platform, introduced years before the generative-AI boom, also helped build an ecosystem around Nvidia hardware.