How do I diagnose which part is making my PC randomly freeze

[stormprince]

Did you change the GPU setting in the BIOS so that the PC "knows" it should use the iGPU instead of the 980Ti you had installed before? This should be done with the 980Ti still installed in the PC.

Back when i had an APU system, i did this when i sold my GPU and started using the APU's iGPU instead.

No i didn't Okay, I'll reinstall it now and try and find the setting

 

[Chomiq]

You could always check Event Viewer in Windows to look for any errors reported at the time the crash happened. You never know, it might show something that will point you in the right direction.

 

[HTC]

No i didn't Okay, I'll reinstall it now and try and find the setting

It works that way with the APU i used but this is an Intel system and i'm assuming it works the same way but, since the last time i owned an Intel CPU was before they came with iGPUs (E8400 ... yeah: that long ago ...), i may be wrong here.

If it does work the same way my APU did, simply "telling" the BIOS to use the iGPU instead of the 980Ti should do the trick. This may be even possible without removing the 980Ti, but i'm not 100% sure: perhaps someone more knowledgeable can help clarify.

You could always check Event Viewer in Windows to look for any errors reported at the time the crash happened. You never know, it might show something that will point you in the right direction.

Also worth a try.

 

[stormprince]

Okay, I've just reseated my graphics card and I just had a idea that when it all started was around the time I moved, also pretty close to that data I installed my CPU watercooler and for
clearance had to take my GPU out, I'm running timespy and seeing if I get greater stability at stock and if I I'll play some games and see if it bricks it

Okay so I put it back in (that what she said) and I think it may have just been seated wrongly as I heard a click that I don't think I heard last time...................
So I ran a stress test from 3D mark and got 98.7 stability at default (earlier at default I got 94-96.6 max) after a few runs and I couldn't get it to increase even whilst underclocking

I'm going to play some games and see if it crashes. I've never overclocked this card (heavily anyway) and I think it's about time I try assuming I run crash-free for a week or maybe less if i'm feeling impatient.
Any tips on overclocking? and specifically checking if the OC I've applied is stable

 

[John Naylor]

1. First thing in these instances is check Windows System Logs in Event Viewer looking specifically at the time stamp for when the crashes are ocurring. Anything that says "warning" (yellow icon) or "error" (red icon) should be addressed.

2, A stress test individually on CPU and GPU shouls be performed.

- CPU We recommend RoG Real Bench which is a suite of "real" applications run in a multitasking environment which tests you system to a far tougher scenario than it will ever encounter. I typically run it for two hours about once ever 6 month tho for a new OC, 4 hours

- For GPU testing, we stick with nVidias recommendation that they suggested some years back when testing laptops and use Furmark ... just wanna see what max temp is ... on creen curve will level out and this is a test to determine viability of cooling system. For atability, run the carious Unigine and 3D mark GFX tests

- For voltage and temp monitoring would recommend HWiNFO

Id ya take the card outta the slot, wipe the card's pins down with 90+% isprophyl alchol ... $2.64 at Walmart and try and clean the slot witha foam Q-Tip the same way. Yes, that click is the lock.... when removing the card many folks forget to press the unlock lever and yank the slot out of the MoBo because they didn't press lever 1st.

 

[Vayra86]

I've taken the 980ti out, and now I can't get a picture.
When I boot the system I get this white glow which goes across the screen and actually shows a faded Xtreme overclocking utility which I had up just before I shut my computer off to take the GPU out and after that, I get a black screen with white glow at the edges.
I've got the monitors lead plugged into DVI, is ridiculous to say my monitor or cable could be an issue right?

To be fair its not that ridiculous. Can you try a new / guaranteed to be good cable?

EDIT: nvm I read you're already on a current test. Go on with that one first. The cable suggestion still stands if the problem persists.

Okay so I put it back in (that what she said) and I think it may have just been seated wrongly

Imagine if she said the second half of that sentence

 

[stormprince]

1. First thing in these instances is check Windows System Logs in Event Viewer looking specifically at the time stamp for when the crashes are ocurring. Anything that says "warning" (yellow icon) or "error" (red icon) should be addressed.

2, A stress test individually on CPU and GPU shouls be performed.

- CPU We recommend RoG Real Bench which is a suite of "real" applications run in a multitasking environment which tests you system to a far tougher scenario than it will ever encounter. I typically run it for two hours about once ever 6 month tho for a new OC, 4 hours

- For GPU testing, we stick with nVidias recommendation that they suggested some years back when testing laptops and use Furmark ... just wanna see what max temp is ... on creen curve will level out and this is a test to determine viability of cooling system. For atability, run the carious Unigine and 3D mark GFX tests

- For voltage and temp monitoring would recommend HWiNFO

Id ya take the card outta the slot, wipe the card's pins down with 90+% isprophyl alchol ... $2.64 at Walmart and try and clean the slot witha foam Q-Tip the same way. Yes, that click is the lock.... when removing the card many folks forget to press the unlock lever and yank the slot out of the MoBo because they didn't press lever 1st.

I'll try furmark, I'm getting higher stability scores now but I'm still experiencing the odd crash, in the event viewer it's saying may not be gigabyte .. I can copy the exact log later

 

[lexluthermiester]

Issue:
After a few hours or occasionally after one, a black, green, orange or purple screen will just appear (a bit like when you've pushed your overclock too high) and I'll have to reboot, sometimes I'll remain in discord and everything will carry on working fine although instant disconnect to the game I was playing.

I was going to suggest re-seating your video card but then you said..

Okay, I've just reseated my graphics card and I just had a idea that when it all started was around the time I moved

So how are things going now?

 

[stormprince]

I was going to suggest re-seating your video card but then you said..

So how are things going now?

I reseated my video card and found that my stability score on 3D mark jumped up way higher, my FPS in games also increases by a significant margin (around 30 fps on apex).
This week I've encountered two crashes (keep in mind this is only a few hours of playing as I'm studying) and they've just been a black or purple screen and I've had to restart.

My temperatures are below 60, and I honestly have no idea why it's crashing. Someone suggested that I stress test my CPU so I think that'll be a good start for me.

 

[lexluthermiester]

I reseated my video card and found that my stability score on 3D mark jumped up way higher, my FPS in games also increases by a significant margin (around 30 fps on apex).
This week I've encountered two crashes (keep in mind this is only a few hours of playing as I'm studying) and they've just been a black or purple screen and I've had to restart.

My temperatures are below 60, and I honestly have no idea why it's crashing. Someone suggested that I stress test my CPU so I think that'll be a good start for me.

I hate to say this, but your card might just be dying..

 

[stormprince]

I hate to say this, but your card might just be dying..

Yeah, the idea crossed my mind, but it's just weird that It does work for hours and hours and then just dies sometimes

 

[Vayra86]

If the card VRAM is already running at stock, you may want to increase the card power target to 110% or so, run at stock core clock, and clock the memory down ~100 in Afterburner.

If the card is on its last legs (degradation), this may prolong its life a bit. If it doesn't make a difference, I'm afraid too that its about to go, probably some sort of VRAM corruption.

 

[lexluthermiester]

If the card VRAM is already running at stock, you may want to increase the card power target to 110% or so, run at stock core clock, and clock the memory down ~100 in Afterburner.

This might work. Give it a shot and let us know how it goes.

 

[aQi]

My specs:
- i7-4790k
- 980ti XTreme gaming waterforce
- 16GB DDR3 Corsair Vengence pro RAM
- 200GB Sandisk SSD 550mb/s
- MSI gaming 5 motherboard Z97
- 750W Corsair bronze power supply

Issue:
After a few hours or occasionally after one, a black, green, orange or purple screen will just appear (a bit like when you've pushed your overclock too high) and I'll have to reboot, sometimes I'll remain in discord and everything will carry on working fine although instant disconnect to the game I was playing.

My attempts:
Running no overclock on anything, reinstalling windows, trying different harddrive, newest and older drivers on GPU only.
I can also run AIDA 64 GPU test for 12 hours without this issue happening once..

Question:
How can I find out, definitely what part is causing the issues? I know it sounds like a graphics card issue but I'm unsure.

Its the ram and im positive about it.

But be sure you are on the latest version of the bios..

 

[lexluthermiester]

Its the ram and im positive about it.

Unless you can see the system in question in person and can troubleshoot on site, you can not be positive. You can do no better than the rest of us, offer hypothesis based on experience and logic.

 

[aQi]

Unless you can see the system in question in person and can troubleshoot on site, you can not be positive. You can do no better than the rest of us, offer hypothesis based on experience and logic.

Im talking clearly on the basis of experience, as i often come across issues like this especially if its corsair used. This isnt the first time as the problem explained is similar to those i encountered, i too was confused and started to switch gfx and hdd and ultimately it was corsair dominators which i had to replace with bunch of normal ddr3 1333mhz rams starting off with 2gb modules, installed windows, driver only and stress tests.
I am positive enough so thats why i am asking him to definitely go for ram replacement after updating the bios.
There was another issue but that was with old systems with NB/SB lga775 but for systems like these iv seen where ram propose such issues.
So just take it easy and let him atleast try with different rams (system bios settings:default)

 

[lexluthermiester]

Im talking clearly on the basis of experience, as i often come across issues like this especially if its corsair used. This isnt the first time as the problem explained is similar to those i encountered, i too was confused and started to switch gfx and hdd and ultimately it was corsair dominators which i had to replace with bunch of normal ddr3 1333mhz rams starting off with 2gb modules, installed windows, driver only and stress tests.
I am positive enough so thats why i am asking him to definitely go for ram replacement after updating the bios.
There was another issue but that was with old systems with NB/SB lga775 but for systems like these iv seen where ram propose such issues.
So just take it easy and let him atleast try with different rams (system bios settings:default)

And I could recite a long list of experiences I've had that would strongly suggest that the video card itself is to blame. The point was neither of us can be certain unless we have direct physical access to the machine in question. Don't let your pride react. It's all theory until the problem is solved and the culprit proven.

 

[aQi]

And I could recite a long list of experiences I've had that would strongly suggest that the video card itself is to blame. The point was neither of us can be certain unless we have direct physical access to the machine in question. Don't let your pride react. It's all theory until the problem is solved and the culprit proven.

Lol. Ok calm down let him evaluate then.

 

[stormprince]

Its the ram and im positive about it.

But be sure you are on the latest version of the bios..

How do I ensure I'm on the latest version of the bios? and should i take one stick out? and see if I experience a crash then take the other out if I do and if that fails just buy some more

Also in the event viewer after a crash i got this:
The description for Event ID 1 from source SMBCtrl cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

Miss BDR table, maybe not GIGABYTE motherboard.
Though I've had crashes in the past where this error was nowhere to be seen

Also some other errors:
Unable to open the Server service performance object. The first four bytes (DWORD) of the Data section contains the status code.
Warnings:
The Open procedure for service ".NETFramework" in DLL "C:\WINDOWS\system32\mscoree.dll" failed with error code Access is denied.. Performance data for this service will not be available.
The Open procedure for service "ESENT" in DLL "C:\WINDOWS\system32\esentprf.dll" failed with error code Access is denied.. Performance data for this service will not be available.
The Open procedure for service "WmiApRpl" in DLL "C:\WINDOWS\system32\wbem\wmiaprpl.dll" failed with error code The device is not ready.. Performance data for this service will not be available.
The Open procedure for service "Lsa" in DLL "C:\Windows\System32\Secur32.dll" failed with error code Access is denied.. Performance data for this service will not be available.
These errors and warnings are repeated throughout the history. Every time there is a crash these appear.

The description for Event ID 1 from source SMBCtrl cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

Miss BDR table, maybe not GIGABYTE motherboard.

That was a different time.

Also, I just got a crash whilst playing counter-strike, which is odd as I've never crashed on that before

 

[aQi]

How do I ensure I'm on the latest version of the bios? and should i take one stick out? and see if I experience a crash then take the other out if I do and if that fails just buy some more

Also in the event viewer after a crash i got this:
The description for Event ID 1 from source SMBCtrl cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

Miss BDR table, maybe not GIGABYTE motherboard.
Though I've had crashes in the past where this error was nowhere to be seen

Also some other errors:
Unable to open the Server service performance object. The first four bytes (DWORD) of the Data section contains the status code.
Warnings:
The Open procedure for service ".NETFramework" in DLL "C:\WINDOWS\system32\mscoree.dll" failed with error code Access is denied.. Performance data for this service will not be available.
The Open procedure for service "ESENT" in DLL "C:\WINDOWS\system32\esentprf.dll" failed with error code Access is denied.. Performance data for this service will not be available.
The Open procedure for service "WmiApRpl" in DLL "C:\WINDOWS\system32\wbem\wmiaprpl.dll" failed with error code The device is not ready.. Performance data for this service will not be available.
The Open procedure for service "Lsa" in DLL "C:\Windows\System32\Secur32.dll" failed with error code Access is denied.. Performance data for this service will not be available.
These errors and warnings are repeated throughout the history. Every time there is a crash these appear.

The description for Event ID 1 from source SMBCtrl cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

Miss BDR table, maybe not GIGABYTE motherboard.

That was a different time.

Also, I just got a crash whilst playing counter-strike, which is odd as I've never crashed on that before

Well the SMBCtrl error is not a new case coming from Gigabyte RBG fusion software. Most of the people reported this and had to uninstall the software.

The second error i presume is from the hard drive or the ram.

For instance can you borrow a value select dim of any brand running at stock speed 1333 or 1600mhz. 2gb will do, see if it installs windows and provides stable running conditions under loads and tests.

But a question i may ask if you are using drivers from gigabyte with RGB Fusion or not ?

I presume you have already tried other HDD for making sure your SSD is not the issue

 

[stormprince]

Well the SMBCtrl error is not a new case coming from Gigabyte RBG fusion software. Most of the people reported this and had to uninstall the software.

The second error i presume is from the hard drive or the ram.

For instance can you borrow a value select dim of any brand running at stock speed 1333 or 1600mhz. 2gb will do, see if it installs windows and provides stable running conditions under loads and tests.

But a question i may ask if you are using drivers from gigabyte with RGB Fusion or not ?

I presume you have already tried other HDD for making sure your SSD is not the issue

Yeah my SSD actually died mid testing so I got a new one and I used a HDD and it still happened whilst I was waiting for my new SSD to arrive.
I can't really borrow any, I can check if my old PC was DDR3 but I really doubt it (it was pentium or celeron)
I actually did have RGB fusion installed I uninstalled it about a hour ago, though the thats the program, how do I ensure I'm not using those drivers?

 

[phanbuey]

Do you have the aorus engine installed by any chance?

 

[aQi]

Yeah my SSD actually died mid testing so I got a new one and I used a HDD and it still happened whilst I was waiting for my new SSD to arrive.
I can't really borrow any, I can check if my old PC was DDR3 but I really doubt it (it was pentium or celeron)
I actually did have RGB fusion installed I uninstalled it about a hour ago, though the thats the program, how do I ensure I'm not using those drivers?

Did uninstalling RGB fusion fixed the SMCrtl for you ?

You should least borrow a ram without wasting money.

Yet you can still try by installing a single dim you have and try fresh installation of windows. Download latest drivers from nvidia. But the SMCtrl was from RGB fusion software not the drivers. But worth trying rather then installing anything from gigabyte just to make sure.

 

[stormprince]

Did uninstalling RGB fusion fixed the SMCrtl for you ?

You should least borrow a ram without wasting money.

Yet you can still try by installing a single dim you have and try fresh installation of windows. Download latest drivers from nvidia. But the SMCtrl was from RGB fusion software not the drivers. But worth trying rather then installing anything from gigabyte just to make sure.

I haven't had a crash since the last one so I can't tell you yet, when I crash I'll look in the event viewer. There was this new error i found though:

Faulting application name: UTHardwareMonitorCmd.exe, version: 0.0.0.0, time stamp: 0x5bdac27d
Faulting module name: KERNELBASE.dll, version: 10.0.17763.475, time stamp: 0x9877ee8b
Exception code: 0xe0434352
Fault offset: 0x0011c762
Faulting process ID: 0x3914
Faulting application start time: 0x01d512653470a935
Faulting application path: C:\Program Files (x86)\GIGABYTE\RGBFusion\UTHardwareMonitorCmd\UTHardwareMonitorCmd.exe
Faulting module path: C:\WINDOWS\System32\KERNELBASE.dll
Report ID: 2595c00b-fd00-4bb4-ac38-9b14a8a55837
Faulting package full name:
Faulting package-relative application ID:]

This did not cause a crash.

 

[aQi]

I haven't had a crash since the last one so I can't tell you yet, when I crash I'll look in the event viewer. There was this new error i found though:

Faulting application name: UTHardwareMonitorCmd.exe, version: 0.0.0.0, time stamp: 0x5bdac27d
Faulting module name: KERNELBASE.dll, version: 10.0.17763.475, time stamp: 0x9877ee8b
Exception code: 0xe0434352
Fault offset: 0x0011c762
Faulting process ID: 0x3914
Faulting application start time: 0x01d512653470a935
Faulting application path: C:\Program Files (x86)\GIGABYTE\RGBFusion\UTHardwareMonitorCmd\UTHardwareMonitorCmd.exe
Faulting module path: C:\WINDOWS\System32\KERNELBASE.dll
Report ID: 2595c00b-fd00-4bb4-ac38-9b14a8a55837
Faulting package full name:
Faulting package-relative application ID:]

This did not cause a crash.

Since you have uninstalled RGB fusion however most of the entries still remain. Use some sort of advance uninstaller to completely wipe off the fusion software.
If you haven’t had a crash may be this was the hickup. Do install drivers from nvidia.com elimnating gigabyte goodies next time.
Tell us where you have reached after that.
Also which version of windows you are using ?