Like HowStuffWorks on Facebook!

How AGP Works


Get Off the PCI Bus
PCI slots on a motherboard
PCI slots on a motherboard
Photo courtesy

In 1996, Intel introduced AGP as a more efficient way to deliver the streaming video and real-time-rendered 3-D graphics that were becoming more prevalent in all aspects of computing. Previously, the standard method of delivery was the Peripheral Component Interconnect (PCI) bus. The PCI bus is a path used to deliver information from the graphics card to the central processing unit (CPU). A bus allows multiple packets of information from different sources to travel down one path simultaneously. Information from the graphics card travels through the bus along with any other information that is coming from a device connected to the PCI. When all the information arrives at the CPU, it has to wait in line to get time with the CPU.

This system worked well for many years, but eventually the PCI bus became a little long in the tooth. The Internet and most software were more and more graphically oriented, and the demands of the graphics card needed priority over all other PCI devices.

Typical example of an AGP-based graphics card
Typical example of an AGP-based graphics card

AGP is based on the design of the PCI bus; but unlike a bus, it provides a dedicated point-to-point connection from the graphics card to the CPU. With a clear path to the CPU and system memory, AGP provides a much faster, more efficient way for your computer to get the information it needs to render complex graphics. In the next section, we'll see how this is done.