Having a bizarre PCIe lane problem

Hi all - I can’t seem to get my graphics cards to use more than 4 PCIe lanes.

I’m running a 12700F on a DDR5 Z690-A ASUS Prime motherboard with the latest version of Windows 11 - I’ve replaced just about everything that could be replaced (motherboard, GPU, ram isn’t the issue obvs), and the problem has persisted - every GPU I use in the system shows up as x4. I’ve switched every bifurcation setting back and forth in the BIOS, cleaned slots with air and solvents, and nothing has made a difference.

At this point, I’m wondering if it’s the CPU itself - has anyone heard of alder lake warping leading to a loss of PCI lanes? All of my NVME drives work, etc., and the only way you’d notice a difference is in benchmarks and a few lost FPS here and there.

I guess I’m looking for a sanity check - would updating my BIOS or reinstalling windows maybe snap something to life? I really don’t wanna pour more money into a dead platform, lol.

Either way, let me know if you all need more information on my end.

Tested cards include an RX 6800 and a Radeon Vega 56 and a B760-CSM mobo aside from the Z690.

Z690 only has one x16 slot. All other slots are either x4 or x1 and often chipset lanes, not CPU.

Check manual and block diagram on your board and you will see.

There is no problem. You just assumed you have more than you really have.

how do you check how many lanes are used :smiley: ?
my GPU is connected to Pcie 4.0 8x Lanes and that is in Powersave @ Pcie 2.0 8x Lanes

Intel®12th Gen Processors*

1 x PCIe 5.0/4.0/3.0 x16 slot(s)

Intel® Z690 Chipset**

1 x PCIe 3.0 x16 slot (supports x4 mode)

1 x PCIe 3.0 x4 slot

2 x PCIe 3.0 x1 slots

is your GPU in the Top or bottom Slot?

1 Like

Yeah that’s a fact, powersave is a thing. But he was talking about “GPUs” plural, so I went with lanes being the problem. Could be both :slight_smile:

i think he switch them for testing the same behaviour :smiley: lets see who is right? ;D

Yup, did the GPU-Z render test to see if it was in powersaving mode, and no changes. This behavior was consistent across the 6800 and the Vega 56.

edit: identical reporting in HWinfo, fwiw

is your GPU in the Top or bottom Slot?

top slot, as would be expected. i’ve also tried multiple NVME drive configurations to ensure that the CPU-specific lanes aren’t being taken up, and no dice.

Ok, right, now we ruled out basic user error.

Yes it is possible that the CPU is not seated correctly. did you hear from those brackets to even them out?

This is why I have a feeling it’s to do with the warping of my chip’s edges - it lived inside a SFF B660 machine for a few months before I decided I wanted a giant unfeeling monolith computer instead of a tiny unfeeling monolith computer, and it’s pretty wavy at this point.

Does anyone have a LGA 1700 pinout handy?

i mean one of those, i had massive thermal probs with my 12th gen 12900k and this bracket solved them

Would it straighten an already warped chip though? It’s been years of warping at this point lol

when the warp is not i say “that” much visble though i would give it a shot you get those brackets realy cheap and what are you going to do next? i gues 13/14th gen? The bracket is still compatible.

Is you Chip older than your warranty by Intel? You could allways try that.

i quick checked 3 Years for Boxed CPUs

It is within RMA period still, I think. i guess we’ll see if this counts enough to be returned lol

You just need to formulate correctly, in your case you dont even have to lie.

Just say your CPU does not provide the correct PCIe Lanes you changed the Board to rule that out so your CPU must be faulty.

Just stay that vague and let them handle the rest. I had good RMA experiences with Intel.

Intel knows they f*ed up the substrate on 12/13th gen :smiley:

1 Like

Clear Cmos, clean socket. I’ll assume you’re not just using the wrong slot from the chipset.

Yeah, I’ll have to give CMOS clearing a shot, and I’ll update the bios too. Gonna see if the less-secure pads on the underside of it have to do with PCIe lanes…

Wait, same problem with same CPU and all GPUs on 2 different boards? Was CPU dropped at some point? Definitely alcohol scrub the LGA pads.

Not to my knowledge, but I’ll definitely tear it all apart and see if i can’t change something in any direction lol

looking at a pinout diagram for LGA 1700, it really does look like the area that I have the most severe warp has to do with the PCIe x16 slots on this corner of the substrate… Maybe I can get lucky and have it reseat properly, but looks like I might need to get a new chip…