Today Nvidia celebrates 25 years of the GeForce 256. This video card was released on October 11, 1999 as the first GPU. This term referred to the single chip that could take care of several 3D related functions.
Nvidia marketing the GeForce 256 as the first GPU that could process at least 10 million polygons per second. It was also the first single 3D chip to offer hardware support for ‘transformation and lighting’, a type of calculation previously done by the CPU. The ‘256’ refers to the 256-bit QuadPipe Rendering Engine, which consists of four 64-bit pixel pipes.
The GeForce 256 had the NV10 chip, which was made on TSMC’s 220nm process. The NV10 had 17 million transistors under the hood and was originally paired with SDR memory. On December 13, 1999, a faster DDR version was released, which increased the bandwidth from 2.65GB/s to 4.8GB/s. For reference, the RTX 4090, Nvidia’s current flagship model, has 76.3 billion transistors and has a bandwidth of 1008GB/s.
Update, 6:32 p.m.: Side note: according to Nvidia’s marketing, the GeForce 256 is the first GPU, although there were similar products from manufacturers such as 3dfx before. The definition of this term is explained at the top of this article.
2024-10-11 15:02:00
#Nvidia #released #GeForce #GPU #years #update