Graphic Processing Unit

Read Complete Research Material



Graphic Processing Unit

The early 1980s have generally been credited with being the roots of the modern era of Computer Graphics. In 1981 IBM created the first video cards available for the PC. The monochrome display adapter was monochrome text only and could not address a single pixel, but had to change the screen in a 9 x 14 pixel area. Although these cards were very limited and only supported non-colored text, at the time they were a good solution considering the limited abilities of the original PC (Fatahalian, 91).

As video cards evolved adding the ability to perform with higher resolutions, greater color depths, and the ability to control individual pixels termed "All Point Addressable", they were still very limited as they relied on the main system CPU to perform all of the work. These new video capabilities only increased the workload on the CPU.

In 1984, IBM created the first processor based video card for the PC. The Professional Graphics Adapter (PGA) incorporated its own Intel 8088 microprocessor onboard and took over all video related tasks freeing the main system CPU from having to perform any video processing computation. This was a very important step in graphical processing unit (GPU) evolution since it helped to further the paradigm of having a separate processor perform computations of the graphics computing system family. The PGA video card required 3 card slots in an XT or AT system, a special professional graphics monitor, and at over $4000 was very expensive (Fallon, 25).

This all added up to a short life span for the PGA card, although to its credit it was targeted at the high-end business market for engineering and scientific applications and not at end users and was capable of 60 frames of 3D animation per second. With the introduction of VisiOn in 1983, the first graphical user interface (GUI) for the PC, larger system requirements and more computational power were prerequisites for the changing and more important role of the graphics computing system family.

New industry wide standards and concepts for graphics computation were required and Silicon Graphics Inc. (SGI) would step in to provide the next stepping stone in the evolution of the GPU. Silicon Graphics Inc. (SGI) was an up and coming computer graphics hardware company that focused their resources on creating the highest performance computer graphics computers available. Along with their hardware they were also creating industry standards that are the basis for today's computer graphics hardware and software. One of the software standards of today is the platform-independent graphics API OpenGL. Created by SGI in 1989, OpenGL has become the industry's most widely used and supported 2D and 3D application programming interface (API) . Another concept (that will be covered later) pioneered by SGI that is now an intricate part of the design of graphics hardware is the idea of a graphics pipeline (Buck, 62). This design approach is nearly the same in all of the major graphical processing unit (GPU) manufacturers like nVIDIA, ATI, Matrox, etc., and ...