Nvidia designed the CUDA parallel-computing platform used by many Nvidia GPUs.
CUDA, first released in 2007, lets developers use Nvidia graphics processors for general-purpose computing rather than only displaying graphics. It provides programming tools, libraries, and application interfaces for workloads such as artificial intelligence, scientific simulation, and video processing.
Nvidia’s market value grew sharply as cloud companies and researchers adopted GPUs for machine learning. The company was founded in 1993 by Jensen Huang, Chris Malachowsky, and Curtis Priem, and its headquarters are in Santa Clara, California.
CUDA is sometimes confused with a type of processor, but it is a software platform and programming model. Nvidia’s competing hardware rivals, including AMD and Intel, provide different software ecosystems.