NVIDIA CORP (NVDA)

NVIDIA is a data center scale AI infrastructure company that pioneered accelerated computing. Its technology stack is built on the CUDA development platform, which runs on all NVIDIA GPUs, plus hundreds of domain-specific libraries, frameworks, SDKs and APIs. These support workloads such as AI model training and inference, data analytics, scientific computing, robotics and 3D graphics. The company takes a platform approach, combining chips, networking, systems, software, algorithms, AI models and services on a unified programmable architecture. Its inventions include the GPU (1999), CUDA (2006), the Tensor Core GPU (2017), autonomous driving SoCs (2018), the DPU following the 2020 Mellanox acquisition, the Grace data center CPU (2023) and the Blackwell architecture. The Rubin platform is expected to begin production shipments in the second half of fiscal year 2027.

NVIDIA reports two segments. Compute & Networking includes Data Center accelerated computing and networking platforms, AI solutions and software, and Automotive platforms and autonomous and electric vehicle solutions. Graphics includes GeForce GPUs for gaming and PCs and Quadro/NVIDIA RTX GPUs for enterprise workstations. It addresses four markets. Data Center offerings include rack-scale systems, GPUs, CPUs, NVLink, InfiniBand and Ethernet networking, and software such as NVIDIA AI Enterprise and vGPU, which are sold under paid licenses. Gaming covers GeForce RTX GPUs, the GeForce NOW cloud service, and SoCs and development services for game consoles. Professional Visualization covers RTX PRO GPUs and Omniverse. Automotive covers the DRIVE Hyperion platform, DRIVE AGX hardware and DRIVE OS. The company also licenses its intellectual property.

Customers include all major public and private cloud service providers, AI model makers, enterprises, startups and public sector entities. Industries served include enterprise software, transportation, healthcare and financial services, along with researchers, gamers, creators, designers and automotive makers. Direct customers include OEMs, ODMs, system integrators and distributors. The company sells through a global partner network that also includes ISVs, add-in board manufacturers and tier-1 automotive suppliers. Application engineers and solution architects provide pre-sales design support. The company reports more than 7.5 million CUDA developers, support for 6,000 applications, and that its GPUs and networking power over 78% of the TOP500 supercomputers. Desktop gaming revenue is typically stronger in the second half of the fiscal year.

NVIDIA is headquartered in Santa Clara, California, and was incorporated in California in 1993 and reincorporated in Delaware in 1998. It is fabless and relies on contract manufacturing. Wafers come from foundries such as TSMC and Samsung, and memory from SK Hynix, Micron and Samsung. Packaging uses CoWoS technology, and assembly and testing are done by Hon Hai, Wistron and Fabrinet. The supply chain is mainly in Asia, with expansion into the U.S. and Latin America. The market is intensely competitive. Competitors include AMD, Huawei and Intel, large cloud companies such as Alibaba, Alphabet, Amazon and Baidu that design their own AI hardware, and providers of high-performance interconnect products.