Nvidia owns the semiconductor-computing platform CUDA. CUDA is Nvidia’s parallel-computing platform and programming model, allowing software developers to use compatible graphics-processing units for general-purpose computation.
Nvidia introduced CUDA in 2006. It expanded the role of GPUs beyond rendering graphics by giving researchers and engineers tools to run highly parallel workloads, including scientific simulations, data processing, and machine-learning applications.
CUDA helped build an ecosystem around Nvidia hardware because developers could use libraries, tools, and frameworks optimized for the platform. This software advantage complemented Nvidia’s chip designs and contributed to its position in data-center and artificial-intelligence markets.
CUDA is not a stock ticker, a separate public company, or a processor family. Nvidia is the corporate owner, while GPUs are the hardware products that execute CUDA-enabled workloads. That distinction matters when comparing Nvidia with chip designers and semiconductor manufacturers.