Defined by Intel in 1990 to standardize communication between the integrated circuits of its chipsets, the PCI standard became a bus extension specification in 1992/93 under the leadership of a consortium regrouping the major manufacturers.(PCI-SIG).
Starting from 1994, with the introduction of version 2.0, the PCI bus quickly replaces the ISA bus.
Basically its a 32-bit extended bus originally clocked at 33 MHz, then to 66 MHz in its PCI pro version. The PCI bus operates a multiple address lines and data, reducing the number of unnecessary contacts between the connectors and expansion cards: the arrangement 32 lines are also used alternately as address bus and data bus. The cost of manufacturing the cards has been sharply reduced.
The main problem of this bus, in its pure PCI version, is that the bandwidth is shared among all elements connected to the bus, when too many cards are installed, it decrease the performance of the bus.
The different PCI bus
gras> PCI 1.0:</gras>the first versiom
PCI 2.0 and 2.1: Out of the specifications of PCI-SIG consortium, it was widespread until the late 1990s.
PCI 2.2 32 bits bus at 33 MHz (as above, a maximum bandwidth of 133 Mb/s) is the most common, still present on the current motherboards.
PCI-X or pro: 64-bit bus to 133 MHz (a bandwidth up to 1066 MB/s), used mainly in the workstations and servers.
PCI-X 2.0: an evolution of the first PCI pro.
Mini PCI: This is the PCI 2.2 in a shortened format to be integrated into laptops.
PCI, is still evolving, a version 3.0 exists, for workstations.
This document entitled « What is PCI » from CCM (ccm.net) is made available under the Creative Commons license. You can copy, modify copies of this page, under the conditions stipulated by the license, as this note appears clearly.