[SOLVED] What's the difference between PCI-E Ports?

[rochismo]

Hello guys, i have finally solved my problem of my GTX 1660 TI underperforming. I had it in the wrong PCI-E Slot (PCI-E x16 2.0 @ 4x 2.0). I have been reading and it doesn't influence that much in performance (theorically) but to me it supposed a huge boost of FPS. Basically playing the witcher 3 on the wrong PCI-E port, full ultra (no hairworks) it was around 45-50 FPS and when i switched it up it boosted up to 70+ FPS everything ultra + hairworks.

I know that the main difference is the data transfer speed or something like that (correct me if i am wrong), but what i don't understand is why people says that there''s no such performance boost when to me there clearly was one huge performance boost.

 

[PC Tailor]

I have been reading and it doesn't influence that much in performance

This is generally between x16 and x8 slots, not x4 - x4 would definitely have a noticeable impact.
The difference is the theoretical maximum bandwidth from the number of lanes that the slot can use:

	PCI-e 1.0	PCI-e 2.x	PCI-e 3.0	PCI-e 4.x
x1	250MB/s	500MB/s	985MB/s	1969MB/s
x4	1000MB/s	2000MB/s	3940MB/s	7876MB/s
x8	2000MB/s	4000MB/s	7880MB/s	15752MB/s
x16	4000MB/s	8000MB/s	15760MB/s	31504MB/s

The reason being as most GPUs don't necessarily utilise more than x8, so the difference is minimal, but cutting the x8 in half would definitely have an impact.

 

[Metal Messiah.]

PCI-E is actually backwards compatible, meaning even if your Motherboard supports gen 2.0, any gen 3.0 GPU will also work on that same board.

There shouldn't be any huge performance drop.

All PCI Express versions are backward and forward compatible, meaning no matter what version the PCIe card or your motherboard supports, they should work together, at least at a minimum level. One important thing to know, however, is that to get the increased bandwidth (which usually equates to the greatest performance), you'll want to choose the highest PCIe version that your motherboard supports and choose the largest PCIe size that will fit.

For example, a PCIe 3.0 x16 video card will give you the greatest performance, but only if your motherboard also supports PCIe 3.0 and has a free PCIe x16 slot. If your motherboard only supports PCIe 2.0, the card will only work up to that supported speed (e.g., 64 Gbit/s in the x16 slot).

The only slight disadvantage is that it will only have the maximum bandwidth provided by the slot; i.e., if you install an x16 video card in an x4 slot, it will have only x4 bandwidth available.

To reach the maximum performance possible, both the expansion card and the PCI Express controller (available inside the CPU or inside the motherboard chipset, depending on your system) have to be of the same revision.

If you have a PCI Express 2.0 video card and install it on a system with a PCI Express 3.0 controller, you will be limited to the PCI Express 2.0 bandwidth. The same video card installed on an old system with a PCI Express 1.0 controller will be limited to the PCI Express 1.0 bandwidth.

 

[rochismo]

This is generally between x16 and x8 slots, not x4 - x4 would definitely have a noticeable impact.
The difference is the theoretical maximum bandwidth from the number of lanes that the slot can use:

	PCI-e 1.0	PCI-e 2.x	PCI-e 3.0	PCI-e 4.x
x1	250MB/s	500MB/s	985MB/s	1969MB/s
x4	1000MB/s	2000MB/s	3940MB/s	7876MB/s
x8	2000MB/s	4000MB/s	7880MB/s	15752MB/s
x16	4000MB/s	8000MB/s	15760MB/s	31504MB/s

The reason being as most GPUs don't necessarily utilise more than x8, so the difference is minimal, but cutting the x8 in half would definitely have an impact.

I've seen videos saying that there's no huge performance from x4 to x16, with an RX 580 of course

 

[PC Tailor]

I've seen videos saying that there's no huge performance from x4 to x16, with an RX 580 of course

What videos?
It vastly varies on what games they are playing and the rest of their system, in modern applications, x4 can definitely be a limitation.

It won't be in every application, for example CPU intensive applications rather than GPU, as the GPU doesn't need the excess bandwidth to process its frames.

 

[BFG-9000]

The impressive 1660Ti is about as powerful as a GTX 980! And at least with the 980, going from PCIe 3.0 x16 to PCIe 2.0 x4 is a 13% average drop in framerate:

I also found Witcher 3 results for the regular 1080:

 

[kerberos_20]

if u use bottom slots for gpu, it will usualy be much slower as they have pcie lanes from chipset which is shared with other peripherals (ssd/network/etc), so u get more throttled then with first pcie slot which got exclusive pcie to cpu lanes