Negative GPU usage and computer freezing

[originalClown]

Hi,

 

I've spoken to Nvidia and AMD support over the last couple of months with no solution. Basically, my computer freezes once or twice a month with -1% GPU usage.

Has anyone ever seen anything like this? No idea what's causing it but it completely freezes up my PC.

GPU is a Nvidia GTX 1070 Founder's Edition and I repasted it last year (Temps seem completely normal).

All drivers are up to date and I've reseated the GPU many times. 

CPU is a Ryzen 7 1800X and is overclocked to 3.9GHz.

RAM is Corsair Vengeance RGB 16GB x2 @3000MHz 

 

I've attached some screenshots of the task manager when a crash occures. If anyone has seen this before or has any suggestions that would be a great help

[Radium_Angel]

10 minutes ago, originalClown said:

overclocked

Turn that, and XMP off, see what happens

[originalClown]

7 minutes ago, Radium_Angel said:

Turn that, and XMP off, see what happens

Yeah I'll give that a go, can't hurt. Never caused an issue before though. Had my CPU and Ram overclocked since I built my PC in 2018. Got an AIO Cooler on my CPU

[Radium_Angel]

1 minute ago, originalClown said:

Yeah I'll give that a go, can't hurt. Never caused an issue before though. Had my CPU and Ram overclocked since I built my PC in 2018. Got an AIO Cooler on my CPU

But running at 110% all the time can eventually cause issues. When troubleshooting things, it's always a good idea to start from stock values, and go from there.

[Matazal]

what specs do you have, I am having the same problem

 

have you seen this post?

 

 

My Computer Specs:

Ryzen 5 3600

MSI RTX 2060 MSI ventus XS OC 6g tu104 Video Card

Corsair 32GB 3200 RAM RGB

Curcial 500GB SSD

WD 2TB HDD

Corsair 650W power supply

MSI B550M Mortor Wifi Mother Board

NZXT 400i Case (with Smart Device v2)

Windows 10 Pro

XMP enabled...

[originalClown]

On 1/12/2021 at 1:40 PM, Matazal said:

what specs do you have, I am having the same problem

 

have you seen this post?

 

 

My Computer Specs:

Ryzen 5 3600

MSI RTX 2060 MSI ventus XS OC 6g tu104 Video Card

Corsair 32GB 3200 RAM RGB

Curcial 500GB SSD

WD 2TB HDD

Corsair 650W power supply

MSI B550M Mortor Wifi Mother Board

NZXT 400i Case (with Smart Device v2)

Windows 10 Pro

XMP enabled...

Yeah just read the thread. I have NZXT CAM running for my Kraken X62 and my PC just froze again today. I've been running without an overclock since I created this thread but clearly it's not making any difference. I seem to get crashes more rarely than the post you shared so no idea what the cause is there. I'll see if that thread's solution changes anything

[Apocalypsing]

I'm aware this topic hasn't been active in a while now, but I thought I'd also chime in as I've also experienced the same symptoms, not just on one computer, but across three of them over time, all with different hardware configurations. I've also had no luck in tracking down the actual cause, in fact it seems nobody has had luck in doing so from my past searching online, given the elusiveness of this issue. I've stumbled across this thread from my search of "negative GPU usage in Task Manager" as that's among the most clear warning signs of this issue from my own experience.

My first run in with this issue began shortly after I ditched Windows 10 LTSC (build 1809), for the regular release channel of Windows 10 (version 2004 to be specific). Version 2004 had been out for several months, I figured by this point it was at least "production ready" in the everyday home use sense, as I had the assumption that a lot of the early growing pains weren't of any concern by that point.

After this new Windows install, everything worked very well initially. Then this issue cropped up for the first time. My description of the problem is what I'd describe as a cascading failure of the OS. From all my computer use over the years, it's not unusual to see the odd glitch here and there. Usually these "glitches" are very minor, and have no real effect on system stability nor performance. But, as I've found, this cascading failure of the OS always starts with a single odd moment. As time progresses, the system only grows more unstable, with more system functionality seemingly falling over, and by the time I've noticed the really odd behaviour, it's often too late to really delve into it, and I may have no choice but to hard reset the system to force restart it.

If I leave the system running instead of hard resetting it, more and more system components in Windows will fall over and become unresponsive over time, whether it's shell components like the start menu, taskbar, or other system applications like task manager failing to start (or close if I've left them open). Often by this point I cannot even open or close other applications. And if I try to unplug any peripherals to try and ease things by this point, said peripherals - whether it's my keyboard, mouse, or audio interface - won't even receive any power from the same USB ports they were plugged into prior. And if I power cycle my monitor, I've often seen a complete loss of signal to the display.

What doesn't help is that despite all of this chaos, with half of Windows 10 falling over, I've never ever seen the OS give me a blue screen during these events, on any of the three computers I've observed it happening on. If I at least had a blue screen I'd have a much better clue as to what's happening. But since I'm having to hard reset to fix these random, soft freezes, the only evidence of something unexpected happening is being logged in Windows Event Viewer. Sadly Event Viewer only sees the event as an unexpected shut down of the OS, and nothing more.

Until this evening, I was convinced my problem stemmed from the two most recent, regular release channel versions of Windows 10 (2004 and 20H2). But tonight I experienced this exact same problem on my 1809 LTSC install for the first time. Naturally, so I've now had to rule out release channel Windows 10 instability, along with other variables I've also ruled out, such as:

 

• Faulty memory: I've done many passes of Memtest86, both on my main 32GB kit as well as my old 16GB kit of memory I still have lying around which I also tested in my main system. No memory errors ever appeared in these tests.
• Nvidia GPU or driver woes: I rolled back to a previous, known good driver version I deliberately kept around from a recent driver update at the time. Also no change, but one of the affected computers doesn't even run on a dedicated GPU, but an Intel integrated GPU, so this was a quick one to rule out.
• Power supply swaps: saw no change here. But again, multiple computers affected, I quickly ruled this one out.
• Swapping out, and avoiding use of certain peripherals: no improvement here either.
• Something something mains electricity in my house: this one may seem far fetched, but it did cross my mind at some point if my issue possibly stemmed from any power stability or quality issues in the house. But my brother's system was one of those affected systems, and he doesn't even live with me. Plus the power grid is well maintained and very stable where I live.
• Software running at the time of instability: when I first ran into this problem it often occurred when I was surfing the Internet, particularly when I was watching YouTube or such. I thought Chrome may have been a culprit at first, so I switched to Firefox full time but saw no improvement from that. My mind then turned to Firefox, as I often had both Firefox and Chrome running simultaneously in the leadup to this problem. Firefox wasn't the issue either. I tested an additional dozen or so applications I also suspected, ruling out each one by one, and ultimately making no progress there either.

Ultimately, I know this reply is a real long one. I feel I've just about exhausted all my options for troubleshooting this, and I'm hoping that by documenting my own experiences in detail here, that other people will find it helpful in grasping a better understanding of what's been ruled out here, even if there's no concrete cause and solution to the issue at present.

Edited April 8, 2021 by Apocalypsing

[mo44]

For what it's worth (in keeping this thread alive) I have this issue too.

Started last year with my 1070 and never found the cause.. then I upgraded to a 6800 in February which wasn't getting this problem until a couple of weeks ago.

 

For me my PC becomes unresponsive randomly at least once a week (especially if I leave it on for multiple days)

 

The only lead I have at this point is that the last AMD driver update appeared to change the PCI Bus driver..

 

Also should add I've not used CAM in many years so it's not solely to blame.

[zeKxD]

Trying to further the search for the root cause.

 

I recently built my new pc back in May (maybe earlier June) and am having the same issues as described. My hardware is almost identical to a couple of posts and has been doin me a concern. But like Apoc, I have also ran extensive tests on every component in my pc and even swapped out some parts and still no avail. 
 

Parts:

CPU: Ryzen 7 3800X

CPU Cooler: NZXT Kraken x73

GPU: MSI RTX 2060

MOBO: MSI MPG B550 Gaming Plus

PSU: Corsair 750M

MEM: 4x8 GB G.skill trident Z RGB (3600)

STORAGE: ADATA SWORDFISH 2TB (swapped out for 500 GB HDD bc I thought it was the cheap m.2)

 

I was running 2 Lian Li PSU extensions, the 24 pin and the 8 pin PCI. Reddit suggested removing those and just using the stock cables that came with the PSU. No change. I was also using the Phanteks vertical GPU mount with a riser cable that came with it, swapped that out and just put the GPU in like a normal build and still no fix. Swapped my m.2 lanes in case it was a lane issue and that didn’t work. Completely removed the m.2 and swapped with a SATA 500 GB HDD and even that didn’t help. A couple of Reddit users were suggesting it was NZXT CAM or MSI afterburner software, but the build worked flawlessly for a couple of months after completion and those 2 softwares are always first thing I do after installing windows updates. If anyone has any clues or suggestions please drop them. 
 

help me obi-wan kenobi, you’re my only hope.

 

[originalClown]

So I haven't had a freeze since January purely by closing NZXT CAM every time I turn on my computer. So I am assuming that this was the issue as I am still running all my overclocks with no problems. 

Hope this helps

[mo44]

On 8/4/2021 at 1:31 AM, zeKxD said:

Trying to further the search for the root cause.

 

I recently built my new pc back in May (maybe earlier June) and am having the same issues as described. My hardware is almost identical to a couple of posts and has been doin me a concern. But like Apoc, I have also ran extensive tests on every component in my pc and even swapped out some parts and still no avail. 
 

Parts:

CPU: Ryzen 7 3800X

CPU Cooler: NZXT Kraken x73

GPU: MSI RTX 2060

MOBO: MSI MPG B550 Gaming Plus

PSU: Corsair 750M

MEM: 4x8 GB G.skill trident Z RGB (3600)

STORAGE: ADATA SWORDFISH 2TB (swapped out for 500 GB HDD bc I thought it was the cheap m.2)

 

I was running 2 Lian Li PSU extensions, the 24 pin and the 8 pin PCI. Reddit suggested removing those and just using the stock cables that came with the PSU. No change. I was also using the Phanteks vertical GPU mount with a riser cable that came with it, swapped that out and just put the GPU in like a normal build and still no fix. Swapped my m.2 lanes in case it was a lane issue and that didn’t work. Completely removed the m.2 and swapped with a SATA 500 GB HDD and even that didn’t help. A couple of Reddit users were suggesting it was NZXT CAM or MSI afterburner software, but the build worked flawlessly for a couple of months after completion and those 2 softwares are always first thing I do after installing windows updates. If anyone has any clues or suggestions please drop them. 
 

help me obi-wan kenobi, you’re my only hope.

 

Problem stopped for me after going back to an old driver, specifically Radeon Software version 21.3.1 (the one before the previously mentioned PCI Bus driver change) so there's something more complicated going on here than it just being CAM.

That solution obviously doesn't help anyone with an nVidia card, which in my case even had the issue long before I got an AMD card. What I do see though is that some people with an nVidia card have had success by disabling Global C State Control in the firmware so that could be worth a try (though that didn't work for me).

[Juli337]

I had the same issue a couple times while watching videos. Gpu at 30°C with copy usage 100% but video encode and decode usage at -1%. Don't know what the problem could be. Maybe something to do with ryzen CPUs? I think no one on this post had an intel cpu. I don't really mind bc it happens rarely and I'm going to change my GPU when prices drop

System parts:

CPU: Ryzen 7 5800x

GPU: EVGA Gtx 750 sc 

MOBO: Asus TUF b550 Plus WIFI

RAM: Corsair Vengance RGB Pro 3200Mhz 2x8GB

PSU: Corsair RM850

Storage: Samsung 970 Evo Plus 2TB

[efadfa]

I've created an account because I struggle with the same issue and it's driving me insane. It often happens when I leave my pc running for the night which is the only time my limited internet allows me to download big updates... Anyway, I have an Intel CPU, to respond to the previous comment. i7 4790K. My computer guy has told me the issue is my old faulty HDD, making the system fall apart. So I put all of my stuff into a new drive and for like a month or 6 weeks it was fine. But it started happening again. At this point I feel we've all been cursed. 

[lordpantsington]

Firefox 93.0, 5900x-6800XT
Plagued by this for the previous 2 days, can be specifically triggered by opening multiple things playing video (twitch, yt, nbcsportsgold).  Turned off hardware acceleration in FF and have not had an issue thus far.  Easy thing to try, unsure if it will last.

[mo44]

I was ultimately able to fix the issue by enabling Memory Integrity (VBS)

[newlpn]

I've been having the same problem as you all for a week now. Disabling iCUE as it comes the closest to NZXT software hasn't helped me yet. Currently trying if it helps to turn off hardware acceleration in FireFox. If that doesn't work, I'll try enabling Memory Integrity. Would be interested if anyone of you has found another solution.

 

[2020Tech4U]

So far the only common thing amongst all of us is AMD Ryzen. I have a 5800X and a red devil 5700 XT and any game 
APP Freezes but when I open TM -1% GPU and restart power off no worky....I've started games with no other apps open and still this drama. 

The only other common thing might be steam and big mode with AMD? Let me know if you are all using Steam Big Mode and AMD.

 

Okay so I have been trying to diagnose the issue and starting with the most basic setup, BIOS settings back to day 1, shutting off any apps that auto open at stat up etc... No tuning No OC's. Well...It actually started right when I got into Windows lol Did not have to load into a game nothing, just entered WIN lol.

So I did system restore back to just a few days to fix that issue. I'll keep posting until I find the culprit. I feel it is something with Steam AMD and WIN updates not playing well with each other. If anyone has anything they'd like to add that hasn't or has worked please reply here so we can found a fix, or Yell at whatever app needs IT on it NOW!

Edited November 2, 2021 by 2020Tech4U
update

[Madzriotzz]

Having the same issue with my 5800x, gtx 1070, and gtx 1080. All on a rog strix b550. I have the asus software installed that has been giving me issues before but this just popped up about a week ago...

[solarbike]

I too am having this issue. but I do not have the new systems you do. My gear was fine (Dell i630 DualDuo something built in 2011 with 8G of memory, 750W supply and originally build in the Vista era but shipped with XP. Paid for W7 and upgraded to 10. Needed a better GPU than GT9800 for a game and got the Asus RTX1030 OC edition. Not a big spender here.

get -1% GPU and lockups once a week. tried everything. sent the GPU back to Asus and they said it was fine and updated the firmware. Problem persists.