GIGABYTE GTX 1060 WindForce 2X OC running at 1x PCI E speed.

[ThisIsMyDog]

My GPU works in PCI-Express 3.0 x1 mode even under full load when playing so the performance is much less than it should be.

I've updated the BIOS, cleaned the PCI Express port and pulled out and re-inserted the graphic card several times.
I also changed some settings in UEFI of MOBO, nothing help.

Rest of PC:
MOBO: ASRock H110M HDV
CPU: Intel Core I5 6500
RAM: 2x4GB A-DATA 2133MHz
PSU: ANTEC VP 500w

 

[rtwjunkie]

Hmmm, I was going to suggest that you change PCIe slots, but you only have one x16 slot.

Did it work correctly before?

Can you uninstall the driver and reinstall a different version, either older or newer? I say older in the hope you had a known good one before.

 

[Assimilator]

Did you check Windows' power saving settings as the tooltip suggests?

 

[ThisIsMyDog]

Hmmm, I was going to suggest that you change PCIe slots, but you only have one x16 slot.

Did it work correctly before?

Can you uninstall the driver and reinstall a different version, either older or newer? I say older in the hope you had a known good one before.

It was working correctly before, i made tests with like 10 different drivers, always the same.

Did you check Windows' power saving settings as the tooltip suggests?

Yes i check these settings, its ok there.

 

[rtwjunkie]

It was working correctly before, i made tests with like 10 different drivers, always the same.

So you are saying that even going back to a driver that it used to work correctly with doesn’t work now?

 

[ThisIsMyDog]

So you are saying that even going back to a driver that it used to work correctly with doesn’t work now?

unfortunately, Yes.

 

[W1zzard]

Try reseating the card, check for damage near the connector area

 

[_UV_]

I've updated the BIOS, cleaned the PCI Express port and pulled out and re-inserted the graphic card several times.
I also changed some settings in UEFI of MOBO, nothing help.

This done after.

My GPU works in PCI-Express 3.0 x1 mode even under full load when playing so the performance is much less than it should be.

What was done before it happens.

check for damage near the connector area

thinking the same

 

[Durvelle27]

Try cleaning the contacts on the connector and spray the slot with some compressed air

 

[ThisIsMyDog]

It was simple cleaning with compressed air and replacement of thermal paste on CPU, then fresh install of system. Could it be damage mobo?

 

[Assimilator]

It was simple cleaning with compressed air and replacement of thermal paste on CPU, then fresh install of system. Could it be damage mobo?

Damaged PCIe slot, possibly. Did you do anything to the PC before this started, i.e. cleaning or something? Or did it just happen for no apparent reason?

 

[ThisIsMyDog]

Sorry i forgot to use quote in previous post.

JUST BEFORE IT STARTED, I cleaned the computer from the dust, replaced the thermal paste on CPU and reinstalled the system.

Now i checked if there are physical damages on MOBO or GPU, I did not see any

 

[John Naylor]

Sounds like the card was no re-seated properly .... did you wipe the card contacts down with alcohol or contact cleaner ? I clean the slots with a foam swab and wipe off the card contacts; in some cases, had to take a pencil easure to the greenish corrosion that can occur in humid environments and lower cost cards / boards.

 

[ThisIsMyDog]

Sounds like the card was no re-seated properly ....

I forgot to mention that when I was cleaning my computer (before this problem appeared) I did not even touch the GPU.

did you wipe the card contacts down with alcohol or contact cleaner ? I clean the slots with a foam swab and wipe off the card contacts; in some cases, had to take a pencil easure to the greenish corrosion that can occur in humid environments and lower cost cards / boards.

I did not do it, when i noticed this problem i only took out the graphic card and cleaned the PCI Express ports with compressed air.


edit/ Now I discovered another problem, the USB data transfer is much slower than before on the same pendrive. Can it be related problems? MOBO do not get enough energy?

 

[_UV_]

JUST BEFORE IT STARTED, I cleaned the computer from the dust, replaced the thermal paste on CPU and reinstalled the system.

main PCI-E x 16 connects to cpu, you probably have damaged/bent some pins inside cpu socket if it was taken out
try reseat cpu

 

[newtekie1]

Check the back of the card at the components I've highlighted. Make sure none are broken or missing.


It could also be the heatsink on the CPU isn't mounted correctly, which is causing the CPU to not make good contact with the socket.

 

[coonbro]

maybe check the card in another capable computer like your buddies gaming rig to see if its the same or at full x16 ? try another good card you know works at full x16 in yours and see if it will or not .

to me it kinda sounds like a power issue not allowing the card to run at its full power ststes and stuck at low power running / x1

use like hwinfo64 to monitor the power usegae at a full load run .. and see

get that info max and min. and check to something like a good review article or here in the gpu data bases .

for the card above

Board power limit Target: 250.0 W Limit: 275.0 W Adj. Range: -40%, +10% [ 250 w X that 10% limit is the 275w max off the slider ]

you have to look up your card core voltage as well to compare like from a review site gpu-z screen shot at load or whatever

one more thing try another / older gpu-z to make sure its not a bug in the one your using under your card . a long shot but could happen

 

[CandymanGR]

Step 1. Uninstall Nvidia drivers from control panel.

Step 2. Restart in safe mode, run DDU (Display Driver Uninstaller) and clean ALL nvidia drivers remnants.

Step 3. Restart normaly. Install the latest nvidia drivers.

Step 4. Check if it works now in x16. Corrupted drivers may cause this.

 

[kastriot]

Check another gpu in your rig or check your gpu in another rig.

 

[eidairaman1]

Check another gpu in your rig or check your gpu in another rig.

Most definitely this.

With problems d
He describes, could be CPU seated badly or cooler too tight

 

[ThisIsMyDog]

maybe check the card in another capable computer like your buddies gaming rig to see if its the same or at full x16 ? try another good card you know works at full x16 in yours and see if it will or not .

to me it kinda sounds like a power issue not allowing the card to run at its full power ststes and stuck at low power running / x1

use like hwinfo64 to monitor the power usegae at a full load run .. and see

get that info max and min. and check to something like a good review article or here in the gpu data bases .

for the card above

Board power limit Target: 250.0 W Limit: 275.0 W Adj. Range: -40%, +10% [ 250 w X that 10% limit is the 275w max off the slider ]

you have to look up your card core voltage as well to compare like from a review site gpu-z screen shot at load or whatever

one more thing try another / older gpu-z to make sure its not a bug in the one your using under your card . a long shot but could happen

Here are my screen from HWiNFO under full load.

http://imgur.com/k4fZc33

Step 1. Uninstall Nvidia drivers from control panel.

Step 2. Restart in safe mode, run DDU (Display Driver Uninstaller) and clean ALL nvidia drivers remnants.

Step 3. Restart normaly. Install the latest nvidia drivers.

Step 4. Check if it works now in x16. Corrupted drivers may cause this.

I did it several times.

Check another gpu in your rig or check your gpu in another rig.

I will be able to do this in a few days.

 

[coonbro]

clocks look in line and show better then speced

Boost: 1797MHz/ Base: 1582MHz in OC Mode

Boost: 1771MHz/ Base: 1556MHz in Gaming Mode
[if your card]
https://www.gigabyte.com/Graphics-Card/GV-N1060WF2OC-6GD-rev-10-11#sp

your gpu power is lower then said , but you may not be running something that will tax it fully as well something like firestrike or maybe heaven bench may them strong clocks at that lesser volts cant be a bad thing then you would think I should atleast hit the target at stock of 120w ? that maybe due to the whatever your running to test with

Board power limit Target: 120.0 W Limit: 140.0 W Adj. Range: -50%, +17%

Memory Clock: 2002 MHz

card looks to be working normal you gopt a vrel with a yes but that comes from your overclock and not enough voltage [core] d it stable may use the3 voltage slider and give it the max allowed or back off the overclock till its clean /no or your not seeing a lot of this in gpu-z

like doing this oc a nice blue field all the way . out of headroom is out of head room can you get more , yes but it will throttle back to

[valley and gpu-z render running at the same time 3ed loop ]

your link speed is fine at 8 so its not like stuck at 2,5 . I don't know if running at pcie x1 you would think It would


id try this if your doing any overclocking like with afterburner or something I would scrap it and uninstall and put the card back to out of the box stock and retest it

 

[newtekie1]

I'll say this once, and once only, this issue is never a software issue.

It might be an issue in the BIOS, some allow you to limit link width. However, this is very unlikely if the OP hasn't been messing around in the BIOS changing things, and will always be fixed by a BIOS reset.

Drivers don't determine link width, power savings don't determine link width. Software/Power Savings will change the PCI-E version from 3.0 down to 1.1, but they will never change the link width down to x1. I'm seriously tired of threads like this coming up and people suggesting a bunch of software fixes. The PCI-E link width is determined during POST and does not change after that.

your link speed is fine at 8 so its not like stuck at 2,5 . I don't know if running at pcie x1 you would think It would

The links speed that HWinfo reports is just the PCI-E version, not the actual speed of the port. 2.5GT/s is PCI-E 1.1 and 8.0GT/s is PCI-E 3.0, both what a PCI-E x1 link provides.