NVIDIA is the semiconductor company that named and popularized the graphics processing unit. Its own corporate timeline records that NVIDIA was “Founded on April 5, 1993, by Jensen Huang, Chris Malachowsky, and Curtis Priem, with a vision to bring 3D graphics to the gaming and multimedia markets.” From the start the company aimed squarely at consumer 3D, a market that in the mid-1990s was crowded with competing accelerator chips and incompatible APIs.
NVIDIA’s early hardware established it as a contender. The RIVA family (Real-time Interactive Video and Animation accelerator) combined 2D, video, and 3D acceleration on a single chip in the late 1990s, and the RIVA TNT line that followed built a reputation as a serious rival in the consumer 3D-graphics market. These parts were fast rasterizers, but like their competitors they still relied on the host CPU to perform geometry transform and lighting.
The pivotal product was the GeForce 256. NVIDIA’s timeline marks 1999 as the year the company invented “the GPU, the graphics processing unit, which sets the stage to reshape the computing industry.” NVIDIA defined a GPU as a single-chip processor that integrated transform, lighting, triangle setup and clipping, and rendering engines, moving the geometry stage off the CPU and onto the graphics chip. The GeForce brand became NVIDIA’s flagship consumer line for the decades that followed.
NVIDIA was also central to the move from fixed-function to programmable graphics. Successive GeForce generations exposed programmable vertex and pixel (fragment) shaders, aligned with Microsoft’s Direct3D shader models and with OpenGL’s shading language. NVIDIA documented these architectures in detail in its GPU Gems books; the GPU Gems 2 chapter on the GeForce 6 Series describes a chip with programmable vertex and fragment processors arranged for high-throughput, single-instruction multiple-data execution.
That same parallel, programmable architecture opened a second act. NVIDIA’s timeline notes that in 2006 the company “Opens parallel processing capabilities of GPUs to science and research with unveiling of CUDA architecture,” letting general-purpose programs run on the GPU rather than only graphics workloads. The deep-learning and AI-training story that grew out of CUDA is covered elsewhere; within the history of graphics hardware, NVIDIA’s lasting contribution is the GPU itself, the rendering pipeline in silicon, and the shift to programmable shading.
This entry uses NVIDIA’s stated April 5, 1993 founding date as recorded on its own corporate timeline.