I have been wondering this ever since ive heard of a graphics card and still havnt figured out can anyone help!!!
**WHAT ARE PIXEL PIPLINES AND WHAT DO THEY DO**
- Thread starter VinnyPel2ella
- Start date
http://www.ati.com/designpartners/media/edudemos/index.html click on 9800, 9700 demo and look at "3d engine" also u can look at other flash demos (like x800)
Your question is pretty vague, and it is difficult to answer it well without actually answering the question "what is a graphics card and what does it do"? 
Basically, the way complex CPUs (and the "GPU"s in modern graphics cards) work is that they break complex operations down into a bunch of simple steps (like an assembly line putting together a car). The output of one step is hooked into the input of the next, and each time the CPU clock is advanced, the data moves from one step to the next. This setup is referred to as a 'pipeline' (the data is 'piped' from one step to the next), and each step is often referred to as a 'stage' in the pipeline.
In a general-purpose CPU, the 'pipeline' is highly programmable, and can be set up to do many different things to the input data (integer or floating point math, write it to memory, or whatever). In a GPU, much of the pipeline is hardwired to do specific things relating to 3D rendering tasks (such as doing color/texture lookups at a particular address, or doing certain matrix transformations) in a particular order, to render a single on-screen pixel -- such a specialized pipeline is called a 'pixel pipeline'. Early graphics cards had one such pipeline, but on modern cards there are up to 24 working in parallel.
In terms of actual results, each pipeline outputs one finished pixel per clock cycle of the GPU. The raw fillrate of a graphics card (in pixels/second) is equal to (number of pixel pipelines * GPU clock speed in Hz). This is the main task done while rendering a 3D scene, and it largely determines the performance of a graphics card in most straightforward rendering situations.
Basically, the way complex CPUs (and the "GPU"s in modern graphics cards) work is that they break complex operations down into a bunch of simple steps (like an assembly line putting together a car). The output of one step is hooked into the input of the next, and each time the CPU clock is advanced, the data moves from one step to the next. This setup is referred to as a 'pipeline' (the data is 'piped' from one step to the next), and each step is often referred to as a 'stage' in the pipeline.
In a general-purpose CPU, the 'pipeline' is highly programmable, and can be set up to do many different things to the input data (integer or floating point math, write it to memory, or whatever). In a GPU, much of the pipeline is hardwired to do specific things relating to 3D rendering tasks (such as doing color/texture lookups at a particular address, or doing certain matrix transformations) in a particular order, to render a single on-screen pixel -- such a specialized pipeline is called a 'pixel pipeline'. Early graphics cards had one such pipeline, but on modern cards there are up to 24 working in parallel.
In terms of actual results, each pipeline outputs one finished pixel per clock cycle of the GPU. The raw fillrate of a graphics card (in pixels/second) is equal to (number of pixel pipelines * GPU clock speed in Hz). This is the main task done while rendering a 3D scene, and it largely determines the performance of a graphics card in most straightforward rendering situations.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter