PCI (Peripheral Component Interconnect), since its inception in 1993, has become the de facto standard input/output (I/O) bus utilized by computer systems. Although computers shipped nowadays also come with an AGP (Accelerated Graphics Port) slot, they still carry the PCI, owing to its widespread use and support from manufacturers.

Definitions

When speaking of Peripheral Component Interconnect, it refers not just to the data that it can carry, but also to the physical features itself, such as its size, protocols, timing and the spacing of the wires and electrical components.

The physical aspects of the PCI bus can take several forms: one of the most common is that of the planar device, which refers to a circuit that is built into the motherboard itself. It can also be an expansion card that can be inserted into a socket. The newest version is the PCI Express.

Functions and Uses

The high speed, coupled with the large data transfer capacity, have turned Peripheral Component Interconnect cards into integral parts of the modern desktop computer or laptop. Among the many functions are sound cards, video cards, TV tuners, modems and those that allow for the use of extra ports for serial and USB devices.

Peripheral Component Interconnect Bus Specifications

The development of the PCI came about due to the need for faster transfer of data; as peripherals and CPU speed increased, the old standard, ISA and VESA, were unable to meet the demands and had to be replaced.

PCI has dual support for 32 bit and 64 bit information data streaming, and it can run, depending on the setup and variation, from 33Mhz to 66Mhz. There are several ways that it can be configured (i.e., when data moves at 32 bits sand speed at 33Mhz, the throughput rate of PCI will reach 133 Mbps). Bus mastering support is also enabled.

Configuration

Unlike its predecessors, it is plug and play, that is, once you have successfully attached the device to the computer, the components will function without the need for any more additional programming or setup.

In most cases, the system will determine the needs of a particular peripheral, and using the PCI Configuration Space, allocate the resources as is necessary. The I/O port will then be divided, its resources allotted to the devices. Besides memory space, ROM can also be allocated, and an x86/PA-RISC code, and an EFI driver.

Instead of utilizing card jumpers, which can result in slow performance, the interrupts (IRQ) are used instead, as it results in faster and better performance. For this reason it is important that your system board has interrupt sharing support, so that the computer bus will know if an interrupt request is for the PCI or for a peripheral on the I/O port.

Historical Development

Work on the Peripheral Component Interconnect began in 1990 with Intel, and two years later version 1 was released, It was version 2, though, which came out in 1993, that witnessed its implementation of standards for the systems/motherboard and the connector itself. Initially it was used in computer servers on networks but it was eventually distributed to desktop PCs, and took over the roles of the old ISA and VESA buses.

Through the 1990s the PCI underwent several upgrades, increasing in clock speed and adding support for form factor signaling.

Even though there are numerous new technologies emerging, it is likely that this component will still be a part of most desktop PCs in the years to come, owing to its compatibility with most existing software.