CertiGuide to A+ (A+ 4 Real)  9  Chapter 9: Graphic Cards

Refresh Rate 1 Understanding 3D Graphics and Functions

Graphics cards fall into one of two general categories - those that simply perform digital to analogue signal conversion, and those that can perform processing of their own. Generally the video cards that just perform digital to analogue conversion are no longer available, and every modern graphics card (even those graphics chipsets subsumed into the motherboard and marketed as "onboard graphics") has its own memory, architecture and processor (referred to as a GPU, or Graphics Processing Unit) that are engineered specifically for graphic related tasks.

These cards are called "accelerated" cards for historical reasons. Back when Windows was first released, the graphical user interface it presented was unheard of. Instead of flat, text based interfaces the user could now drag windows, click icons and use a palette of different colors as they wished. This all required processing power & memory to accomplish, and in those days, the CPU was king of the graphics world. The only trouble was, the CPU in question was an Intel 286MHz processor - if you were lucky. In addition, of course more time spent on manipulating areas of memory for graphic operations meant less time performing "useful" processing. That is not even taking into account the memory required for the graphical display was being taken from the system memory pool.

Manufacturers saw that consumers liked the graphical interface, but did not like the speed penalty. Therefore, to relieve the stress on the CPU they invented the accelerated graphics card. With these cards, instead of the CPU having to allocate an area of system memory and manipulate it, it could pass all the work onto the graphics card. This allowed the CPU to send a single instruction to the graphics card to move the data from one position in memory to another, and then to get on with useful processing whilst the graphics card dealt with the fine details of the move asynchronously.

Over the years, this accelerated technology has evolved to allow incredible graphics processing on desktop machines in the form of 3D acceleration. The consumer games industry is almost entirely responsible for the advances of 3D accelerated graphics cards over the last 7 to 10 years, with progress and demands for greater performance fuelled by games such as Doom, Quake, Half Life and Max Payne. These 3D cards, such as the NVIDIA GeForce range, are specially designed for 3D operations using a sequence of operations called the Transformation and Lighting, or T&L, pipeline. This pipeline is responsible for taking 3D graphic input, and converting it to a 2D representation that can be displayed on a monitor. It also provides all the neat tricks in modern games, such as bump mapping, lighting and vertex/pixel shader effects. By taking the bonus off the CPU for processing the calculations necessary for these effects, the CPU again had more time to work on other tasks, whilst the graphics card could be optimized purely for graphics processing.

Refresh Rate 1 Understanding 3D Graphics and Functions