Question Lag, Crash and Hang! Plagued 2 months with this problem. Please help.

zerocool1431

Honorable
Nov 14, 2015
13
4
10,515
System: i7 6700k
Asus Z270 F gaming
Asus strix R9 390x

Antec EAG 750W Gold

Kingston Hyperx 8gb

Silicon power Nvme 512gb

The Problem: So much of system lag that I feel like banging my head on the desk. Even the mouse pointer lags! Black Screens and eventually a lagged system with no response.
I initially thought it must be a GPU or PSU issue (as most of the above problems have gpu or psu as problems in the countless threads I've come across, maybe 100+) so I RMA'd the PSU and have a new one now and even the SSD.
Things that I've tried with my friends pc:
He owns and i5 7th gen, B250 Mobo, 1060 3gb, 500w Antec Vp series
Things that worked flawlessly:
  1. Swapped out only my gpu and put it in his system
  2. Swapped out his gpu and put it in my system
  3. Swapped CPU, PSU ans every possible combination (was at his place for 10+ hours)

Things that didn't work:
  1. My (Mobo Z270 + GPU R9 390X) + His (CPU i5 7th gen + PSU + HDD + RAM)
Somehow whenever I connect my gpu to my Mobo, the problem always comes back.
So, I took another 1060 but this was a 6Gb version.
That's because I thought maybe the motherboard is having some issues when handling bigger GPUs.
And VOILA! The lag returned again.

As of now I have my own i7 6700k, hyperx 8gb, Asus Z270 Mobo but the only thing different is the GPU which is a 1060 6Gb instead of my R9 390x
And the system still lags.
Is it possible that the mobo cannot handle things whenever a bigger card is used or maybe because the number of pcie lanes used are more by the 6gb 1060 and the 8gb 390x?
I seriously don't know if a motherboard can act this crazy. Please help me with this and tell me if this is possible.
(Yes everytime I've uninstalled the drivers with DDU and even tried older drivers, cleared CMOS, reinstalled windows)
Running the system without a GPU on integrated 6700k's graphics is also smooth without any issues.
The issue only persists when a bigger gpu is attached to the mobo.
Please please please help me with this. I would be very grateful is someone shed a light on this and helped me with things to do ahead. I have a lot of work that has been backlogged due to this issue. Thanks in advance
 
  • Like
Reactions: okjak808

okjak808

Honorable
Feb 23, 2018
227
29
10,840
I think your board has at least 2 PCIE X16 express slots, have you tried both of them?

if the board works without any GPU installed it could be a issue with the slots itself which you can’t fix unless you do some major soldering work, but I assume your board should be in warranty still and you can try RMA it if anything.

Because of that you ran your board with your IGPU and it worked fine just not when you install a GPU.

some motherboard have PCIE lanes that can’t handle graphics cards that use more than 35 watts mostly OEM boards or them mini ITX boards. But your board should be able too so it shouldn’t be an issue and it’s not that the board can’t handle bigger cards or because of the PCIE lanes.

even with a GT 710 which has 8 lanes should still work fine in a 16x lane so that shouldn’t be a problem.
 

zerocool1431

Honorable
Nov 14, 2015
13
4
10,515
I think your board has at least 2 PCIE X16 express slots, have you tried both of them?

if the board works without any GPU installed it could be a issue with the slots itself which you can’t fix unless you do some major soldering work, but I assume your board should be in warranty still and you can try RMA it if anything.

Because of that you ran your board with your IGPU and it worked fine just not when you install a GPU.

some motherboard have PCIE lanes that can’t handle graphics cards that use more than 35 watts mostly OEM boards or them mini ITX boards. But your board should be able too so it shouldn’t be an issue and it’s not that the board can’t handle bigger cards or because of the PCIE lanes.

even with a GT 710 which has 8 lanes should still work fine in a 16x lane so that shouldn’t be a problem.


Missed the warranty by 2 months. Eneded in October 2020.
The top one is x16 and the lower is x8
However the board only seems to give issues when a higher memory gpu is attached to it. (on BOTH slots)
I installed a 1060 3gb and it worked very well without any issues.
I might sound dumb but is it possible that maybe some pcie lanes aren't working which might be causing this amount of lag?
 

okjak808

Honorable
Feb 23, 2018
227
29
10,840
That’s very strange. I had a GTX 1050 4GB in a Emachine system which was like 12 years old and it worked fine. Your motherboard is modern and most of the features that comes with the board are meant for extreme usage like gaming, editing, streaming or other intensive stuff.

So the PCIE seems to work fine on your board but it won’t work properly with a card that uses more than 3gb of vram. My guess is that maybe this has to do with a software issue or drivers within your motherboard, maybe you can try resetting your BIOS back to default to see if that fixes it, or you could try updating your BIOS Firmware, BUT make sure it’s the right BIOS Firmware for your board or else it’ll be bricked making it unusable.

since you already used DDU it’s most likely not a GPU driver issue.

Also how old is your build, was it recently you just built this pc or awhile back, and when did you start experiencing this issue, did you make anything changes to your system, installed a software that could be causing this issue, changed something in your pc that could’ve started this issue?
 
Last edited:

zerocool1431

Honorable
Nov 14, 2015
13
4
10,515
That’s very strange. I had a GTX 1050 4GB in a Emachine system which was like 12 years old and it worked fine. Your motherboard is modern and most of the features that comes with the board are meant for extreme usage like gaming, editing, streaming or other intensive stuff.

So the PCIE seems to work fine on your board but it won’t work properly with a card that uses more than 3gb of vram. My guess is that maybe this has to do with a software issue or drivers within your motherboard, maybe you can try resetting your BIOS back to default to see if that fixes it, or you could try updating your BIOS Firmware, BUT make sure it’s the right BIOS Firmware for your board or else it’ll be bricked making it unusable.

since you already used DDU it’s most likely not a GPU driver issue.

Also how old is your build, was it recently you just built this pc or awhile back, and when did you start experiencing this issue, did you make anything changes to your system, installed a software that could be causing this issue, changed something in your pc that could’ve started this issue?

The board is 3 year old and the processor is 5. I had an Asrock Z170 which i sold and got this one. It has been working completely fine all these years. There was no recent software installed apart from the usual ones I've been using. Adobe Premiere, After Effects, Audition, MSI Afterburner to keep the fan profile in check since its a 390x and I live in a humid environment. This problem appeared out of nowhere around 2-3 months ago when I was using my system just like every other day.
At first I thought it was a driver issue, then i thought it might be some hardware issue, (main suspect was gpu), but that solved out and later I tried windows reinstall for a couple of times. Even tried 2019 drivers which were recommended by people on threads who where facing the same issue. Whenever the lag came in followed by intermittent black screen the drivers used to keep crashing and the event manager used to show that "amdkmdap has restarted".
(This was the reason I tried multiple drivers)
But later the Nvidia 1060 6gb (with proper drivers) gave the same issue so obviously it is not a driver issue or the gpu issue.
BIOS updates are long gone for this board as the last update was on 2018/04/18 and i have been running on that bios for the past 2 years easily.
 

okjak808

Honorable
Feb 23, 2018
227
29
10,840
The error “amdkmdap has restarted” seems to be only happening to AMD GPU’s and as you said it’s mainly a driver issue or either a faulty hardware. In some cases the psu needed to be replaced or it can even be your ram that could be causing this as well.

although since you switched the gtx 1060 6gb what type of errors do you get, is it the same “amdkmdap has restarted” error or is it different l?

take a look at these threads though they’re mainly for amd cards.

https://community.amd.com/t5/graphi...d-responding-and-has-successfully/td-p/277487

https://community.amd.com/t5/driver...dkmdap-stopped-responding-and-has/td-p/100290

View: https://www.reddit.com/r/AMDHelp/comments/8raiq1/ram_causes_display_driver_to_crash_amdkmdap/
 

zerocool1431

Honorable
Nov 14, 2015
13
4
10,515
The error “amdkmdap has restarted” seems to be only happening to AMD GPU’s and as you said it’s mainly a driver issue or either a faulty hardware. In some cases the psu needed to be replaced or it can even be your ram that could be causing this as well.

although since you switched the gtx 1060 6gb what type of errors do you get, is it the same “amdkmdap has restarted” error or is it different l?

take a look at these threads though they’re mainly for amd cards.

https://community.amd.com/t5/graphi...d-responding-and-has-successfully/td-p/277487

https://community.amd.com/t5/driver...dkmdap-stopped-responding-and-has/td-p/100290

View: https://www.reddit.com/r/AMDHelp/comments/8raiq1/ram_causes_display_driver_to_crash_amdkmdap/

Sorry for the late reply. I would like to thank you for doing so much of research and going through the threads.
However I've apparently have already gone through those threads and even tried the drivers they recommended.
The PSU has been recently replaced under warranty and is hardly a month old.
In case of the 1060 the errors are about some service named " nvlddmkm ". I guess that's the service that handles the Nvidia drivers.
I would also like to mention the following thing that I tried: (Most important part) :
My (Motherboard + GPU) + A friend's (Antec 500w PSU + RAM + i5 7th gen + HDD + all the needed things)
And YET the issues persisted. Which is why I took out the PSU and RAM out of the culprit list.

As of now I am writing this post on my own computer but without any gpu in it. (hooked the HDMI staright to mobo so using the integrated 530) and it's running completely fine.
What I usually do it is Run the Heaven Benchmark and run 2-3 forza 1080p 60fps videos with stacked browser windows and keep alt+tabbing while constantly moving the mouse to notice any lag. And still as of now the system ran fine running 2 1080p 60fps videos + benchmark + a few more tabs in mozilla and task manager.
It seems that we have understood where the problem is coming from but we have no idea what exactly is causing it and that is making me pull my hair off.
I was supposed to make documentaries and give it to the people whom i worked for and I'm trying my best to explain them but clients always think that those are just nonsense reasons.

New update:

I have a copy of the movie Pacific Rim 2013 4k 10bit HDR
When I played this movie with the gpu the system seemed like it cried for help and cannot even play one movie in VLC with nothing else running. Playing the movie seemed like a struggle but god f orbid if I skipped it to a later part by clicking on the timeline bar, the system went crazy and hanged as if I am playing this movie on a pentium with some 20 year old GPU..
But right now ? The movie is running perfectly fine on the integrated graphics. :/
 
Last edited:

okjak808

Honorable
Feb 23, 2018
227
29
10,840
Yes that error you got “nvlddmkm” is just like that error you got with your AMD card. Strange because this is due to a driver issue. Although most solved this problem by using DDU, while others had to go through a long process of trying out every single driver update within that card, some had luck others didn’t.



A person also had this issue and it was solved do to one of his ram sticks at fault, but in your case you’ve used your friends ram and the problem still occurred. This could be more than a driver issue but a hardware issue as well which is my theory. Though it’s different for everyone so I’m not too sure.

If you haven’t try using only 1 stick of ram and test it in every slot to see if maybe one of the dimm slots are at fault. Try running sfc/scan in windows CMD, it’s maybe possible that a file in windows could be corrupted.

another solution I found is within the power options in control panel. Go to control panel>power options>Create a power plan>Select “High Performance”, after you done that click on “change power settings” then go to “change advanced power settings”, once you’re there look for “PCI EXPRESS” Tab and click on the plus + sign and turn it off and hopefully this will help.

you can also try this, but be very careful when editing in registry because one mistake can mess up your entire pc, but if you’re up for the risk here you go, but this if from another website.

  1. Goto registry editor (start -> run -> regedit)
  2. Navigate yourself to HKEY_LOCAL_MACHINESystemCurrentControlSetControlGraphicsDrivers (make sure GraphicsDrivers is highlighted).
  3. Right click -> New -> DWORD (32-bit) value -> Name it: TdrDelay -> double click it -> Set it to A (hexadecimal) or 10 (decimal).
here is the website, it may be old but most of process should be similar to modern windows.

https://www.google.com/amp/s/www.instructables.com/How-to-Fix-the-Nvlddmkm-Error/?amp_page=true
 
Last edited:

zerocool1431

Honorable
Nov 14, 2015
13
4
10,515
Yes that error you got “nvlddmkm” is just like that error you got with your AMD card. Strange because this is due to a driver issue. Although most solved this problem by using DDU, while others had to go through a long process of trying out every single driver update within that card, some had luck others didn’t.



A person also had this issue and it was solved do to one of his ram sticks at fault, but in your case you’ve used your friends ram and the problem still occurred. This could be more than a driver issue but a hardware issue as well which is my theory. Though it’s different for everyone so I’m not too sure.

If you haven’t try using only 1 stick of ram and test it in every slot to see if maybe one of the dimm slots are at fault. Try running sfc/scan in windows CMD, it’s maybe possible that a file in windows could be corrupted.

another solution I found is within the power options in control panel. Go to control panel>power options>Create a power plan>Select “High Performance”, after you done that click on “change power settings” then go to “change advanced power settings”, once you’re there look for “PCI EXPRESS” Tab and click on the plus + sign and turn it off and hopefully this will help.

you can also try this, but be very careful when editing in registry because one mistake can mess up your entire pc, but if you’re up for the risk here you go, but this if from another website.

  1. Goto registry editor (start -> run -> regedit)
  2. Navigate yourself to HKEY_LOCAL_MACHINESystemCurrentControlSetControlGraphicsDrivers (make sure GraphicsDrivers is highlighted).
  3. Right click -> New -> DWORD (32-bit) value -> Name it: TdrDelay -> double click it -> Set it to A (hexadecimal) or 10 (decimal).
here is the website, it may be old but most of process should be similar to modern windows.

https://www.google.com/amp/s/www.instructables.com/How-to-Fix-the-Nvlddmkm-Error/?amp_page=true


So,
  1. Tried the ram and everything works fine.
  2. sfc/scannow and repairing has been done along with later followed by reinstallation of windows a couple of times and as I said even connected another HDD (friend's) and the same problem came up with his system. So corrupted windows file is out of the question
  3. ASPM for PCIE devices is been off from the bios as well as windos\
  4. And as far as the TdrDelay is concerned, the issues do persist but it just gives the drivers a few extra time before they decide to crash I guess? Lag does happen but crashing happens after some time.
As the mobo works fine with just a smaller gpu and with the iGPU I'm thinking of selling the board and processor and switching to Ryzen.
I was planning for the 3600 but I guess the prices will be dropped when the 5600 comes up along with the 5300, 5400, 5500 sometime in Jan 2021.
So I don't want to regret buying a 3600 when the 5000 series comes in. (3600 is 245USD here as of now and I'm pretty sure the 5600 or the lower variants would be somewhere along the same price as the 5600x is 380usd)
I would be happy to switch to even a 5400 if that's 6 core 6 thread or even a 4 core 8 thread as that would be a big leap from the 6700k because of the IPC and bigger cache and higher pci lanes despite being the same core count.