PCI Express

(redirected from Arapaho (Computer Bus))

PCI Express

The common hardware interface in PCs, Macs and other computers for connecting peripheral devices such as storage drives and graphics cards. PCI Express (PCIe) was introduced in 2002 as "Third Generation I/O" (3GIO), and by the mid-2000s, motherboards had at least one PCIe slot for graphics. PCIe superseded PCI and PCI-X.

Switched Architecture - Multiple Lanes
Unlike its PCI predecessor, which used a shared bus, PCI Express is a switched architecture of up to 32 independent, serial lanes (x1-x32) that transfer in parallel. Each lane is full duplex (see illustration below).

Internal and External for Laptops
A mini PCIe came out for laptops (see Mini PCI Express) and Thunderbolt extends PCIe outside the computer (see external GPU). For PCIe/PCI comparisons, see PCI-SIG. See PC data buses, PCI, M.2, ExpressCard, Thunderbolt and PCI-X.
PCI Express   Data Rate (each direction)Version   1 Lane  4 Lanes  16 Lanes (MBps)        (GBps)

 Gen 1      250       1        4
 Gen 2      500       2        8
 Gen 3      984       3.94    15.85
 Gen 4     1969       7.88    31.51
 Gen 5     3938      15.75    63.02
 Gen 6**   7877      31.51   126.03 
  ** = planned for 2021

Parallel Transfer of Serial Channels
Each lane is an independent single-bit serial channel. PCIe is a type of parallel transfer but each lane has its own clock, which differentiates it from earlier parallel technologies.

PCIe on the Motherboard
This Asus motherboard has four x1 and three x16 slots (two black, one beige). The x16 slots accommodate x16, x8 and x4 cards. (Image courtesy of ASUStek Computer Inc.)

M.2 Over PCIe
This is a 960GB NVMe SSD on an 80x20mm M.2 card. The connection is via PCIe, either via an M.2 socket on the motherboard or a PCIe adapter card. See M.2.

Different Sizes of PCI Express
PCIe sockets are not the same as PCI, and they come in x1, x4, x8 and x16 sizes.

PCIe Replaced AGP for Graphics
The AGP slot gave way to an x16 PCIe slot for the graphics card. (Image courtesy of NVIDIA Corporation.)
Copyright © 1981-2019 by The Computer Language Company Inc. All Rights reserved. THIS DEFINITION IS FOR PERSONAL USE ONLY. All other reproduction is strictly prohibited without permission from the publisher.