There are masses of different graphic card decisions for purchasers, and to make the very best choice it behooves them to be educated. To that end it is critical to explore the ever expanding feature set found on modern graphic cards.
The best starting point understanding the modern graphic card is in its interface. Where the graphic card plugs into the motherboard describes by Interface and is mostly a PCI-Express for Graphics (PCIE-G) slot unless the graphic card and system in question are reasonably old, in which case the interface will be some flavour of complicated Graphics Port (AGP). PCIE-G interfaces only come in 2 variations : X16 and X16 2.0, while AGP comes in X1, X2, X4, and X8 variations.
In the case of AGP, the bigger the multiplier (the last digit) is, the speedier the communication speed, and so the bigger potential for higher performance graphics. PCIE-G slots only come on in 2 variants, the 2.0 spec being twice as quickly as the standard PCIE-G slot. Again, the quicker the facility to transfer information from the system to the graphic card, the larger the potential performance that they can produce.
Of course, that’s potential and not precise performance. Lower performing graphics cards heading for PCIE-G 2.0 spots won’t be in a position to offer seriously better graphics performance than high-end AGP X8 graphic cards. To define tangible performance one must figure in the memory size and speed, as well as the performance of the graphical processor.
The memory performance comes from both its size and speed. In the MegaBytes (MB) size of graphic card memories are usually measured at the time of this writing, but it appears likely that Gigs will become the dominant standard in the early to mid 2010s. Basically, the more memory a graphic card has, the less often it must make requests for help from the noticeably slower system memory, frequently called RAM. It may help to compare memory to a desk size, and processing jobs to bits of paper all drafted in the same font ; the bigger the job, the more sheets of paper, and the better quality pictures, the bigger those scraps of paper will be. Once the desk has become inundated by paper, one has to find someplace else to put the papers, and swapping them in and off the desk requires time. Thus , a bigger memory pool on cards (or desk) would be perfect.
Naturally, having all of the memory in the world is worthless if the memory is only slightly quicker than the main system memory. Memory clock speeds mixed with how much they read and / or write at the same time (bandwidth) are both vital factors for deciding memory performance of any graphic cards. The quicker the memory and / or the bigger the bandwidth, the better the memory subsystem of the graphic card is probably going to perform. The particular graphics processing core (GPU) will also play a serious factor, both re speed and capabilities. New graphical standards need more recent GPUs that provide support to them.
