Jump to content

Random computer switch offs - GPU? - HDD?

Hey all, Thx in advance for the help!

 

Issue

For a while now my PC will randomly freeze for a few seconds and then so what LOOKS like a restart. It can be under next to no load like searching through web pages or while raytracing spiderman on high details. It seemingly is random I've noticed zero corelation. I could be wrong im no expert but it feels more like a restart that an all out crash (But maybe dont put too much stock into that)

 

When it restarts the screens (x3) all freeze during this time I can still hear my teammates on discord and they can hear me. Then the screens loose their signal as it restarts.
There hasn't been any blue screen error codes.

 

Recent Steps Taken

I have reseated the GPU in its riser and reseated the riser itself, I have also reseated all the ram and updated my drivers. Recently I have also done a wipe of all my drives and re-installed windows. This issue pre-dated that reset however its becoming more and more frequent. (There was a time It felt like it was happening when I bumped the desk but we are talking a seriously light bump which is what lead me to reseat everything. This being said I tired to recreate the effects by diving the desk a bump and ever managed to manually make this happen)

 

PC "Issues" (Listing all known issues I've had incase any of it is more relevant to you guys than me)

GPU - My gpu sometimes shows in windows task manager as having no load on it even in games that I KNOW there is a load on it for sure. EDIT: Also my windows background I've set on all my monitors just goes black sometimes (like just now hence the edit!)

HDD - My HDD is NOT my windows boot drive and stores only some steam games and world of warcraft on it but was causing issues pre reinstall (hence the re-install)

Sata SSD - There is a sata SSD I was trying to use for a while but for some reason refused to show its exsistance in BIOS or windows (Maybe a dead drive)

 

I've attached the CPUZ list of my parts also, along with a few other bit that may or may not prove useful.


Specs:

CPU: AMD Ryzen 7 3700x

MoBo: B450 Tomahawk MAX

RAM: 32Gb DDR4 3200 corsiar vengence rgb pro (4 sticks of 8)

GPU: Gigabyte RTX 3070TI

PSU: 850w ROG Strix

STORAGE:

M.2 - 500gb (Boot Drive)

M.2 Thru PCI Adapter 1TB

HDD 1TB

 

Any insight anyone can provide would be seriously appreciated I'm at such a loss.

 

Thanks so much!

Drives.png

CPU.png

Mobo.png

ram.png

gpu.png

On idle after a restart.png

High Mem processes.png

Edited by Emjack24
Forgot a possable symptom
Link to comment
https://linustechtips.com/topic/1501337-random-computer-switch-offs-gpu-hdd/
Share on other sites

Link to post
Share on other sites

First thing, I spotted is that you're running PCIE 4 on a B450, That BIOS is outdated and could be the cause. I would recommend updating it. As far as I'm aware, the RTX 3070 does not saturate a PCIE 3 x16 link so you should experience identical performance on PCIE 3. Does it only "restart" when you're in a game? I would also test with just 1 stick of ram if the "restarts" still happen.

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

8 minutes ago, JenjerBread said:

First thing, I spotted is that you're running PCIE 4 on a B450, That BIOS is outdated and could be the cause. I would recommend updating it. As far as I'm aware, the RTX 3070 does not saturate a PCIE 3 x16 link so you should experience identical performance on PCIE 3. Does it only "restart" when you're in a game? I would also test with just 1 stick of ram if the "restarts" still happen.

The issues only started a few months ago and there was no hardware changes during that time. I could be wrong but wouldn't my config vs bios have casued these problems from the moment I built the PC?  It doesn't only happen in games and happen when scrolling facebook or watching youtube also.

My friend just pointed out my GPU is running 3.0 x 8 for some reason. GPUZ pic attached

2023-04-17 10_31_39-TechPowerUp GPU-Z 2.52.0.png

Link to post
Share on other sites

23 minutes ago, JenjerBread said:

First thing, I spotted is that you're running PCIE 4 on a B450, That BIOS is outdated and could be the cause. I would recommend updating it. As far as I'm aware, the RTX 3070 does not saturate a PCIE 3 x16 link so you should experience identical performance on PCIE 3. Does it only "restart" when you're in a game? I would also test with just 1 stick of ram if the "restarts" still happen.

Also just spotted this when I streamed my screen on discord for a friend. 75% GPU load

WTH.png

Link to post
Share on other sites

3 minutes ago, Emjack24 said:

The issues only started a few months ago and there was no hardware changes during that time. I could be wrong but wouldn't my config vs bios have casued these problems from the moment I built the PC?  It doesn't only happen in games and happen when scrolling facebook or watching youtube also.

My friend just pointed out my GPU is running 3.0 x 8 for some reason. GPUZ pic attached

2023-04-17 10_31_39-TechPowerUp GPU-Z 2.52.0.png

Usually when GPUs aren't under load or using bandwidth the PCIE link will downclock to save power. I'm pretty sure that's expected behaviour on all modern GPUs. I would still update the BIOS anyway since there will be patches that improve stability. If it doesn't fix it, you can always downgrade back to your current version to keep PCIE 4 support. Do you remember when the issue first started occurring? was it after a windows update?

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

17 minutes ago, JenjerBread said:

First thing, I spotted is that you're running PCIE 4 on a B450, That BIOS is outdated and could be the cause. I would recommend updating it. As far as I'm aware, the RTX 3070 does not saturate a PCIE 3 x16 link so you should experience identical performance on PCIE 3. Does it only "restart" when you're in a game? I would also test with just 1 stick of ram if the "restarts" still happen.

Just want to add that cpuz for some reason says pci-e 4.0 on my board also, its done this thru several updates too, even though its a evga x299 dark and couldnt possibly run that, so that may just be a hiccup in the software and its really doing pci-e 3.0 like mine. It may need a bios update though, i dont know. Overwolf seems to have some conflict with programs, Razer's software is shit and always has been so that may be another issue, Firefox is pretty good but will clobber your ram and cpu cycles a bit, GET RID of geforce experience it will just kill your game settings, gpu driver and ram while spying on you etc...its just shit.

Link to post
Share on other sites

1 minute ago, Emjack24 said:

Also just spotted this when I streamed my screen on discord for a friend. 75% GPU load

WTH.png

that does seem a bit high for discord. Are you sharing at a higher resolution maybe? or it could just be discord not being nice.

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

9 minutes ago, JenjerBread said:

Usually when GPUs aren't under load or using bandwidth the PCIE link will downclock to save power. I'm pretty sure that's expected behaviour on all modern GPUs. I would still update the BIOS anyway since there will be patches that improve stability. If it doesn't fix it, you can always downgrade back to your current version to keep PCIE 4 support. Do you remember when the issue first started occurring? was it after a windows update?

I'll give a bios update a go. Annoyingly i cant make it crash so it could be a day or two before I get a sense of if its worked.

Also FYI I just furmarked the GPU temps where good but the bus interface was still on x8 not x16.

Link to post
Share on other sites

5 minutes ago, Emjack24 said:

I'll give a bios update a go. Annoyingly i cant make it crash so it could be a day or two before I get a sense of if its worked.

Also FYI I just furmarked the GPU temps where good but the bus interface was still on x8 not x16.

not a problem with the time to crash, just update back / reply/quote and ill get a notification. There should be an option in GPUZ to Load the PCIE bus if i recall. Running the Furry Donut doesn't always push the link. Try clicking the question mark next to the GPU link speed that you highlighted / red boxed in your Screenshot. It should bring up a window + option to test link.

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

24 minutes ago, JenjerBread said:

not a problem with the time to crash, just update back / reply/quote and ill get a notification. There should be an option in GPUZ to Load the PCIE bus if i recall. Running the Furry Donut doesn't always push the link. Try clicking the question mark next to the GPU link speed that you highlighted / red boxed in your Screenshot. It should bring up a window + option to test link.

Done that its still x8 3.0 👎

Link to post
Share on other sites

6 minutes ago, Emjack24 said:

Done that its still x8 3.0 👎

Ok, it might be worth reseating again.

What riser are you using? does it support PCIE 4? It could be that the riser included with the case doesn't support the full speed hence the reduced link. I would try using the GPU without the riser if possible.

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

1 minute ago, JenjerBread said:

Ok, it might be worth reseating again.

What riser are you using? does it support PCIE 4? It could be that the riser included with the case doesn't support the full speed hence the reduced link. I would try using the GPU without the riser if possible.

I've just contacted the company about that. It's a GPU riser that came with my Thermaltake p3 core "case" (Chassis). as its not just a riser its made for this case specifically. Unfortunatly I cant find the specs of it so I've emailed them. Could a damaged riser cause theses crashes? and show the reduced lanes?

Link to post
Share on other sites

Just now, Emjack24 said:

I've just contacted the company about that. It's a GPU riser that came with my Thermaltake p3 core "case" (Chassis). as its not just a riser its made for this case specifically. Unfortunatly I cant find the specs of it so I've emailed them. Could a damaged riser cause theses crashes? and show the reduced lanes?

damaged risers can definitely cause reduced lanes. Not too sure about crashes. I would assume that the PC would fail to post if certain wires were shorted. but it could be internal breakage leading to data lines missing and errors be caused from that.

 

I saw that you had a open style case. Could you lay it flat and use the GPU in the PCIE slot directly. should avoid a floating GPU situation (and avoid stress on the PCIE Slot) while diagnosing the issue?


How frequently are the crashes?

I edit my messages often. Please refresh for the latest message.

I am not always on the same thread. Tag or Quote me to get my attention.

Remember to Mark the solution to help others in the future.

 

SFF

CPU: R7 5800x3D

Motherboard: Asus ROG Strix x570-i

RAM: Corsair vengeance 64GB 3200MHz CL16

Cooler: Asetek 645LT

GPU: EVGA GTX 1080Ti Black

Case: Dan Case A4

Storage: 2TB Sabrent (PCI-E 3.0) - Windows

Storage: 1TB Crucial P5 Plus (PCIE 4.0) - Fedora

Storage: 1TB Samsung 860 EVO (SATA) - Steam

 

Link to post
Share on other sites

1 hour ago, JenjerBread said:

damaged risers can definitely cause reduced lanes. Not too sure about crashes. I would assume that the PC would fail to post if certain wires were shorted. but it could be internal breakage leading to data lines missing and errors be caused from that.

 

I saw that you had a open style case. Could you lay it flat and use the GPU in the PCIE slot directly. should avoid a floating GPU situation (and avoid stress on the PCIE Slot) while diagnosing the issue?


How frequently are the crashes?

The reduced lanes may be because the available cpu lanes are being used or have been split because of HDD's,SSD's etc... and maybe x8 is all thats left. It could also be a nearly full HDD or SSD doing what seems like a crash cause it ran out of swap space.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×