Question System instability, intermittent freezing ?

[Nighthawk117]

Hi guys, I'm having an intermittent issue with my pc freezing. Typical symptoms are a sudden loss of control where I can move the mouse but cant open any new programs or shutdown the computer, If I have task manager open it will show the cpu usage graph as frozen. The problem can manifest itself in days or weeks which makes it very difficult to troubleshoot. I've also had the occasional blue screen (these are rare), on the previous one I checked the memory dump and it had memory corruption error. Only the bluescreens seem to leave any kind of trace as to what happened.

The instability is not load dependant, most of the freezing has been on the desktop with light loads, a couple were under heavy loads. Temperatures for both CPU and GPU are kept below 70C under heavy loads so this isn't a cooling issue.

My first thought was a memory issue, I have 4 x 8GB Corsair Vengeance 3000mhz modules. The pc does seem to fail more frequently in this configuration. However I have tried a different 16GB Corsair Vengeance kit from another pc that I know works fine and I still had issues with that.

I tried re-installing Windows 10 on another disk but same symptoms. Any ideas what this is most likely to be? I'm at a bit of a loss, thanks.

My specs are:
Cpu: Core i9 10850K @ 4.5Ghz
Cooler: Asus Strix LC 240mm
Motherboard: Gigabyte Z490 Vision G
Ram: 32GB Corsair Vengeance 3000mhz
GPU: Palit Nvidia 3080
SSD: 2 x 1TB nvme, 2 x 500GB sata
HDD: 2 x 4TB
PSU: Seasonic Focus GX Gold 1000W
Case: Meshify C

 

[Fix_that_Glitch]

Is your chipset driver up to date? IME Driver

 

[SkyNetRising]

My first thought was a memory issue, I have 4 x 8GB Corsair Vengeance 3000mhz modules. The pc does seem to fail more frequently in this configuration.

Can you show screenshots from CPU-Z - memory and spd sections?
(upload to imgur.com and post link)

 

[Nighthawk117]

Is your chipset driver up to date? IME Driver

I've just checked and yes the IME Driver appears to be the latest.

 

[helper800]

I've just checked and yes the IME Driver appears to be the latest.

Have you tried using the shortcut Control + Shift + Windows button + B when the computer is "frozen"? This resets the graphics driver which may temporarily fix the freeze. If this does not fix the freeze there is either something wrong with the win10 install, the storage device its on, or possible still even the graphics card.

Have you tried testing the system with the integrated GPU on the CPU? If the problem is your graphics card this will most likely confirm or deny it if the issue goes away after about 3 weeks to a month.

Have you done an sfc /scannow scan in an admin command prompt? This will rule out anything wrong with the OS install if it comes back clean.

Have you tried running memtest from windows to determine if your RAM kit is showing any issues? This will definitively tell you if the RAM kit is at fault with even a single error reported.

Have you done any mining on the 3080 at all?

 

[Nighthawk117]

Can you show screenshots from CPU-Z - memory and spd sections?
(upload to imgur.com and post link)

Have you tried using the shortcut Control + Shift + Windows button + B when the computer is "frozen"? This resets the graphics driver which may temporarily fix the freeze. If this does not fix the freeze there is either something wrong with the win10 install, the storage device its on, or possible still even the graphics card.

Have you tried testing the system with the integrated GPU on the CPU? If the problem is your graphics card this will most likely confirm or deny it if the issue goes away after about 3 weeks to a month.

Have you done an sfc /scannow scan in an admin command prompt? This will rule out anything wrong with the OS install if it comes back clean.

Have you tried running memtest from windows to determine if your RAM kit is showing any issues? This will definitively tell you if the RAM kit is at fault with even a single error reported.

Have you done any mining on the 3080 at all?

I havn't tried resetting the GPU driver when its frozen, I can perhaps look into that. To try and rule out a problem with the GPU I did put my old 1060 back in and have had the same problem. Also I used the 3080 from Sept 2020 to Feb 2021 on my old 6700K system without issue. I've done no mining on the 3080. I've tried system sfc /scannow and never found an issue. When I upgraded to my current system I didn't reinstall Windows, however in case that was a problem I've recently installed a fresh installation on another SSD but it still has the same issue.

I've tried memtest outside of Windows and that didn't find anything, I've tried running it inside Windows and that didn't show anything either but I didn't run it as long, probably about an hour.

Any ideas how long it would normally take for errors to show up?

Thanks

 

[helper800]

Any ideas how long it would normally take for errors to show up?

Thanks

Errors can happen at any point in the testing.

Have you tried flashing BIOS to the latest version? Have you tried using DDU to uninstall drivers and then installing the independent Nvidia drivers that you can DL from their website?

 

[Nighthawk117]

Errors can happen at any point in the testing.

Any advice on how long it's best to leave it running? Also any preferences for Memtest in Windows vs outside? Thanks

 

[helper800]

Any advice on how long it's best to leave it running? Also any preferences for Memtest in Windows vs outside? Thanks

I would do at least 4 passes in the extended or full test. This can take upwards of 8 hours with 32gb of RAM though.

 

[Nighthawk117]

I would do at least 4 passes in the extended or full test. This can take upwards of 8 hours with 32gb of RAM though.

Ah ok I havn't run it for 8 hours, can I just check which Memtest, would that be:

TechPowerUp

MemTest64 is a free system memory stability and error detection software for all Intel and AMD processors. It runs on 64 bit Windows without installation.

www.techpowerup.com

 

[helper800]

Ah ok I havn't run it for 8 hours, can I just check which Memtest, would that be:

TechPowerUp

MemTest64 is a free system memory stability and error detection software for all Intel and AMD processors. It runs on 64 bit Windows without installation.

www.techpowerup.com

I was referring to the windows memory testing platform, just to be clear.

 

[SkyNetRising]

I've just checked and yes the IME Driver appears to be the latest.

Can you show screenshots, I asked you before?

Can you show screenshots from CPU-Z - memory and spd sections?
(upload to imgur.com and post link)

 

[Nighthawk117]

Can you show screenshots, I asked you before?

Hi, screenshots below:

Intel ME

https://imgur.com/ddTEpSk

View: https://imgur.com/ddTEpSk


CPU-Z Memory & SPD tabs

https://imgur.com/7Tie8sg

View: https://imgur.com/7Tie8sg


Thanks

 

[SkyNetRising]

CPU-Z Memory & SPD tabs

Memory settings look good.
Make sure DDR voltage is set to 1.35V.

You may try to lower ram frequency a bit - to 2800mhz or 2666mhz.

 

[Nighthawk117]

Memory settings look good.
Make sure DDR voltage is set to 1.35V.

You may try to lower ram frequency a bit - to 2800mhz or 2666mhz.

DDR voltage is set to 1.35V, I have tried 2666mhz before but at 1.2v and got the same issue. I'll look into lowering the frequency at 1.35v.

 

[Nighthawk117]

Have you tried using the shortcut Control + Shift + Windows button + B when the computer is "frozen"? This resets the graphics driver which may temporarily fix the freeze. If this does not fix the freeze there is either something wrong with the win10 install, the storage device its on, or possible still even the graphics card.

Have you tried testing the system with the integrated GPU on the CPU? If the problem is your graphics card this will most likely confirm or deny it if the issue goes away after about 3 weeks to a month.

Have you done an sfc /scannow scan in an admin command prompt? This will rule out anything wrong with the OS install if it comes back clean.

Have you tried running memtest from windows to determine if your RAM kit is showing any issues? This will definitively tell you if the RAM kit is at fault with even a single error reported.

Have you done any mining on the 3080 at all?

Just frozen again and I tried the 'Control + Shift + Windows button + B' to reset the GPU driver but no effect. Was worth a try but I've tried two GPU's now anyway.

 

[Nighthawk117]

Anyone ever encountered issues with faulty ram slots on motherboards?

 

[helper800]

Anyone ever encountered issues with faulty ram slots on motherboards?

Yes, I would try the other 2 slots 2b 4b and see if that helps.

 

[Fix_that_Glitch]

It may take you a while to check them all, but you can try using one stick at a time and see if the problem persists or stops with each stick.

 

[Nighthawk117]

Hey guys,

Just for reference, I got to the bottom of this in the end. It was the motherboard, as soon as I replaced it for a different model the instability stopped, all other components including the RAM are fine.