companies

Nvidia

Progettista di chip le cui GPU e il cui stack software, tra cui NIM e DGX Cloud, alimentano la maggior parte dell’addestramento e dell’inferenza dei modelli di IA nel mondo.

In inglese

Nvidia was founded in Santa Clara, California, in 1993 by Jensen Huang, Chris Malachowsky and Curtis Priem, initially as a maker of graphics chips for PC gaming. The company's CUDA programming platform, introduced in 2006, let developers run general-purpose computing on its GPUs rather than just graphics rendering, and that early bet on programmable parallel computing became the foundation for its dominance in AI training roughly a decade later, as deep learning researchers adopted GPUs over CPUs for training neural networks. Since the mid-2010s, Nvidia's data-center business, built around GPU families like the A100, H100 and Blackwell generation chips, has become its primary growth driver, overtaking gaming as the company's largest revenue segment during the generative AI boom. Nvidia reported roughly 42,000 employees in its fiscal 2026 annual report, up sharply from prior years, and became one of the most valuable public companies in the world during the 2023-2026 AI infrastructure buildout. Beyond chips, Nvidia has built out a software and services layer, including NIM (pre-packaged inference microservices), DGX Cloud (managed GPU supercomputing sold through cloud partners), and NVIDIA AI Enterprise licensing, aimed at capturing value from the AI stack beyond raw silicon and deepening customer dependence on its ecosystem.

Founded

1993

Headquarters

Santa Clara, United States

Sector

hardware