Inside a computer, there are special devices that are used to connect units for data and information to flow between two or more devices. These collections of wires are called buses. However, if you really want to get technical, buses that only connect two devices are called ports. Buses are how the various devices within your computer talk and communicate with each other.
In most personal computers, there are two key buses. There is the system bus, or the local bus, which connects the brain of the computer, the central processing unit (CPU), and the system memory. The second key bus is the shared bus, which connects additional components like a modem, a hard drive, and a sound card to the computer.
To increase the speed of your computer, a new type of bus was invented. This is called the Dual Independent Bus, or DIB. DIB broke up the system bus into two separate buses: a frontside bus and a backside bus. The backside bus was specifically used for only one purpose- to connect the CPU to the level 2 cache. The backside bus maximized the speed at which a computer could operate by isolating the cache. Meanwhile, the frontside bus was used to connect the CPU to the system memory and to other buses.
Every bus consists of two important parts: the data bus and the address bus. The data bus transports the actual data and information between two devices. On the other hand, the address bus carries the exact location of where the data is located in the memory or where the data should be stored in the memory.
Buses can be measured by their width, speed, and bandwidth. The width of a bus refers to how wide a bus is (duh!). The wider the bus is, the more information it can carry. Think of the roads you can find in your neighborhood. The small roads can carry fewer cars than the big highways. This is similar to buses-wider buses can carry more information than smaller buses. The speed of a bus is how fast it can transport information. Bus bandwidth refers to how much information a bus can transport in a certain amount of time. It is usually used to measure how many millions of bytes of information a bus can transport from one device to another in one second.
ISA:
ISA is an acronym for Industry Standard Architecture. It is the type of bus found in almost all modern personal computers, and can be used to connect slower peripherals like modems, low-speed networking, as well as some older devices.
PCI:
The Peripheral Component Interconnect, or PCI, is a fairly new bus standard. This bus is used to connect as many as five additional peripherals (mostly important high-speed devices like storage disks, video cards, and high-speed networks) to the computer.
The PCI is connected to the CPU by connecting to the system bus. It connects to the system bus via a bus bridge. Bus bridges are commonly used to put together all of the data from various other buses and then to send all the information to the system bus.
AGP:
The AGP, or Accelerated Graphics Port, is a port (a bus that only connects two devices) that is specifically used to connect graphics cards to the central processing unit and to the system memory.
Because computers rely so heavily on graphics (for example, high-tech computer games, graphs, and other software), people realized that in order to enhance the performance and speed of graphics hardware, they would need to create a special port that would specifically be used for graphics cards. Originally, graphics cards were connected via the PCI, but now PCIs are considered to be slow and bottlenecked. The AGP is much faster than the PCI because it can directly connect to the system memory at full speed and because the only device that uses it is the graphics card. The PCI bus connects many other peripherals, not only the graphics card, to the computer.
Another reason why the AGP is so fast is because it pipelines information. This means that with a single request, the port can receive multiple chunks of data. To get a clear view of how this works, imagine you are at a fancy restaurant and that you are about to order a five-course dinner. You could first order one dish, wait for the waiter bring it to you, then order the next dish, wait for the waiter to bring that to you, and so on. Or, you could order all of the dishes at once, and have the waiter bring them to you as they are ready. The latter description is how the AGP works. As you can see, it's much more efficient than ordering dishes one by one. Furthermore, the AGP port can receive new instructions while still carrying out old ones!
Because of the AGP, computers can have
the easy-to-use graphical user interface, in which users can interact with
their computer by using a mouse and just by pointing and clicking. Also, we
can play games with the high-tech graphics and realistic images. These can
all work quickly and efficiently because of the AGP!
1995-2001, ThinkQuest Inc.all rights reserved
