The video card acts like a "middle man," working between the processor and the monitor. The processor generates the data to be displayed. The video card then translates this information into a form (analog for CRTs and digital for LCDs) that the monitor can display. With older video cards, the main processor did all of the work necessary to generate the data to be displayed. The displayed data was even stored in main system memory. Eventually, memory was added to the video cards, freeing main system memory for other uses.
With the advent of graphical operating systems, such as Windows, much more display-related data had to be generated. The main processor spent more and more of its time computing display data, reducing overall performance. To clear this bottleneck, companies added coprocessors to their video cards, so that most of the necessary video calculations could be done on the card, reducing the work load on the main processor. Video cards with coprocessors are known as video accelerators or graphics accelerators.
Most modern video cards contain a graphics coprocessor, memory, RAMDAC, and a bus interface. They also are provided with software drivers. Together, these components help determine the relative performance of your PC's graphics.
The graphics coprocessor determines what the video card can and can't do. It handles many of the calculations needed to display 2D graphics and video. To make an object or scene more realistic, there are 3D graphics techniques available that generate a graphical representation of the object along three axes (height, width, and depth). These techniques "trick" the PC user into seeing a 3D image on a flat (2D) screen. Specially-designed graphics coprocessors are available for generating the enormous amount of data needed to produce 3D images.
The amount of memory determines the maximum resolution and color depth that your PC can support. The more memory, the more pixels and colors that can be displayed on the monitor. For example, 2.4MB of memory is needed to display true-color (24-bit) graphics at 1024x768 resolution (1024 pixels x 768 pixels x 3 bytes). Double buffering is usually needed when running 3D graphics. This means that the card must be capable of holding two sets of scenes in its memory, thus, twice the memory is needed.
RAMDAC, the card's digital-to-analog converter and color manager, determines how frequently the video card can refresh the image on the monitor. (Note: Monitors need refreshing 60-90 times per second.) It is called RAMDAC because the data comes directly to it from the memory contained on the card. The faster the RAMDAC, the more quickly the card can update the displayed data and the more stable the images seem to appear on the screen. Many graphics coprocessors include a built-in RAMDAC.
With the introduction of the Pentium processor, video cards were designed to plug into the PCI bus, which is a 32-bit wide bus that can operate at data rates of 132 MB per second. With the advent of the Pentium II processor, Intel introduced a new graphics-only bus, called the Accelerated Graphics Port (AGP). This bus can handle data rates up to 526 MB per second. The PCI bus is adequate for 2D graphics and video. However, cards designed for the AGP bus should be used when running programs that exhibit 3D graphics and more sophisticated multimedia applications.
Although software drivers are not part of the video card, they perform a critical function. The drivers enable the video card to talk to and take orders from the operating system and its applications.
To specify a video card's capabilities, video standards were established in the early years of the PC, primarily by IBM. The intention of these standards was to define agreed upon resolutions, colors, etc., to make it easier for the manufacturers of PCs, monitors, and software to ensure that their products work together. One of these standards, the Video Graphics Array (VGA), provides for a resolution of 640x480 pixels with 16 colors. VGA has become the de facto standard for graphics and is the minimum provided on all video cards manufactured today. It is also used by Microsoft when the Windows operating system starts up in the Safe Mode. It was the last well-defined and universally accepted standard for video cards.
With IBM's fall from dominance, no universally accepted hardware standards have emerged. Manufacturers have produced a new generation of extended or super VGA (SVGA) cards that support a variety of different (not necessarily compatible) video modes. The Video Electronics Standards Association (VESA) has been formed to define new video interface standards. The VESA SVGA standard is called the VESA BIOS Extension, or VBE. This standard, now in its second version, can be implemented in either hardware or software. More and more manufacturers are starting to comply with it.
By the late 1990s, there were a number of manufacturers producing highly specialized graphics coprocessor chips for handling the ever expanding 3D graphics and multimedia applications. However, because of the intense competition, there are now only two major players: Canada's ATI Technologies and California-based nVidia. ATI produces both graphics chips and the complete video cards; however, nVidia only produces the chips and sells them to other vendors who build and sell the cards. NVidia's current series of chips are called GeForce.
The type of video card you need depends on the usage. For example, if you use your system primarily for word processing, spreadsheets, and 2D graphics, a video card with 8MB of memory that plugs into the PCI bus is probably adequate. These cards cost in the neighborhood of $50. However, if you're into 2D image processing and 3D graphics/video, a faster card with more memory is needed. Such 3D graphics cards are now available for use with high-end games, computer-aided design, and multimedia applications. These cards come with 16MB to 64MB of memory and utilize the AGP bus. They typically cost in the range of $100-$200.
Manufacturers of video cards include ATI Technologies (www.ati.com), ASUS (www.asus.com), Gainward (www.gainward.com), Hercules (www.hercules.com), Leadtek (www.leadtek.com), PNY Technologies (www.pny.com), and Visiontek (www.visiontek.com).
T. Zinneman is a CCPCUG member. If you have any questions or comments, please send them to tzinneman@comcast.net.
For more information on the Tulsa Computer Society click here