I've been looking at some of the reviews on the HBA Controller cards on OCuK, and have noticed that some peeps still get confused with the various PCIe lanes, so I thought i'd throw together a quick primer.
PCI-Express or PCIe is the supercession for the older PCI and PCI-X interface.
The older technologies were parallel whereas PCIe is serial, meaning for those few who don't know, that it has a single point to point interface or link rather than all the slots sharing one controller.
In the case of PCIe, two slots share one link, which is why when you put two GPU's on your mobo, they scale down the number of useable lanes.
PCIe slots are configured in powers of 2, so you have x1, x2, x4, x8, x16, and x32, although realistically I can't see x32 slots appearing any thime soon, with the largest in common use at the moment being x16. (x referring to the number of lanes)
A PCIe card will fit into a slot of its physical size or bigger.
Some professional boards use open-ended sockets to permit physically longer cards and will negotiate the best available connection.
The number of lanes actually connected to a slot may also be less than the number supported by the physical slot size. An example is a x8 slot that actually only runs at x1; these slots will allow any x1, x2, x4 or x8 card to be used, though only running at the x1 speed.
In practice we only see x1, x4, x8 and x16 on current equiment.
Your motherboard may have this type of socket is described as a "x8 (x4 mode)" slot, meaning it physically accepts up to x8 cards but only runs at x4 speed. The advantage gained is that a larger range of PCIe cards can still be used without requiring the motherboard hardware to support the full transfer rate — in so doing keeping design and manufacturing costs down.
With regards to transfer speeds:
In PCIe 1.x , each lane carries 250 MB/s. (Gen1-signalling mode)
PCIe 2.0, doubles the rate to 500 MB/s. (Gen1-signalling mode)
PCIe 3.0, currently in development (for release around 2010), will add a Gen3-signalling mode, at 1 GB/s.
To give you some idea as a comparison, with parallel interfaces, the quickest PCI slot was a PCI-X 133 which was a 64bit slot.
A single-lane PCIe x1 card has nearly twice as much bandwidth as the most common PCI interface, a 32-bit 33MHz PCI bus (133 MB/s) The one that's on your boards at the moment.
A PCIe x4 slot has bandwidth comparable to the fastest version of PCI-X 1.0
(64-bit 133MHz.) making the PCIe x4 SATA/SAS disc controller cards the same as the best PCI-X controllers, and x8 Controllers exponentially faster.
An eight-lane slot has a transfer rate comparable to the fastest version of AGP.
Hope this has been usefull.
PCI-Express or PCIe is the supercession for the older PCI and PCI-X interface.
The older technologies were parallel whereas PCIe is serial, meaning for those few who don't know, that it has a single point to point interface or link rather than all the slots sharing one controller.
In the case of PCIe, two slots share one link, which is why when you put two GPU's on your mobo, they scale down the number of useable lanes.
PCIe slots are configured in powers of 2, so you have x1, x2, x4, x8, x16, and x32, although realistically I can't see x32 slots appearing any thime soon, with the largest in common use at the moment being x16. (x referring to the number of lanes)
A PCIe card will fit into a slot of its physical size or bigger.
Some professional boards use open-ended sockets to permit physically longer cards and will negotiate the best available connection.
The number of lanes actually connected to a slot may also be less than the number supported by the physical slot size. An example is a x8 slot that actually only runs at x1; these slots will allow any x1, x2, x4 or x8 card to be used, though only running at the x1 speed.
In practice we only see x1, x4, x8 and x16 on current equiment.
Your motherboard may have this type of socket is described as a "x8 (x4 mode)" slot, meaning it physically accepts up to x8 cards but only runs at x4 speed. The advantage gained is that a larger range of PCIe cards can still be used without requiring the motherboard hardware to support the full transfer rate — in so doing keeping design and manufacturing costs down.
With regards to transfer speeds:
In PCIe 1.x , each lane carries 250 MB/s. (Gen1-signalling mode)
PCIe 2.0, doubles the rate to 500 MB/s. (Gen1-signalling mode)
PCIe 3.0, currently in development (for release around 2010), will add a Gen3-signalling mode, at 1 GB/s.
To give you some idea as a comparison, with parallel interfaces, the quickest PCI slot was a PCI-X 133 which was a 64bit slot.
A single-lane PCIe x1 card has nearly twice as much bandwidth as the most common PCI interface, a 32-bit 33MHz PCI bus (133 MB/s) The one that's on your boards at the moment.
A PCIe x4 slot has bandwidth comparable to the fastest version of PCI-X 1.0
(64-bit 133MHz.) making the PCIe x4 SATA/SAS disc controller cards the same as the best PCI-X controllers, and x8 Controllers exponentially faster.
An eight-lane slot has a transfer rate comparable to the fastest version of AGP.
Hope this has been usefull.
Last edited:
