August 31, 1999 nVIDIA (as the name was then spelled) announced the GeForce 256 graphics card, which was released on October 11 of that year. Advertised as the «world’s first GPU», the video card completely changed the market. The term GPU (short for Graphics Processing Unit) was also coined by NVIDIA.
A revolutionary change in the GeForce 256 was the addition of the transformation and lighting (T&L) function directly in the GPU. Previously, these calculations were processed by the CPU or separate hardware. By integrating T&L into the GPU, NVIDIA has reduced costs and improved performance over traditional 3D accelerators. With the architectural leap and optimized software, NVIDIA was able to outperform the competition.
According to the company’s co-founder Chris Malachowski, NVIDIA had 35 competitors at the time of the GeForce 256 release. Some companies were acquired by NVIDIA, others left the market — only ATI survived, which was acquired by AMD. Today, the GPU market includes Intel, AMD, and Qualcomm, but NVIDIA dominates. Since the GeForce 256, GPUs have improved significantly, now equipped with video encoding units, ray tracing, and tensor cores. They are crucial in gaming, professional visualization, cloud computing, and artificial intelligence.
The GeForce 256 had SDR memory, but was later upgraded to DDR, and the video card supported DirectX 7.0. The chip had 23 million transistors — in 25 years, their number has increased 3300 times in GeForce RTX 4090and the memory capacity has increased 768 times.
Source: VideoCardz