New parts causing reboots

Associate
Joined
9 Nov 2003
Posts
429
Location
nr Stoke
I recently got a 3080FE which showed my aging X99 system up so replaced the CPU, motherboard and boot drive and since it has been built I have been suffering with fairly frequent restarts (4 times today). These only appear to happen when I am browsing or watching a video. I haven't had a single restart while gaming/benchmarking.

My specs are as follows:

Ryzen 5600x
Gigabyte B550 Vision D-P
3080FE
32GB (4x8GB) Corsair DDR4 2400mhz (will eventually be replaced but not until this is fixed)
WD SN850 2TB
Seasonic 760w Platinum PSU

I'm at a bit of a loss as to what would be causing it as no errors are thrown or reported in the Event Viewer, PC just restarts. 3080 was ran for a few weeks within my old 5820k system without issue so don't believe it is that. RAM, while not the ideal speed, was fine in the old system and have tried with just 2 sticks but still getting reboots. Also passed whatever Microsoft's own memory test is called without any errors. Motherboard is running the latest bios and have updated all my drivers to the latest I can find.

I did think I had the issue sussed when the 8 pin CPU power seemed to not be fully inserted as it ran for a couple of days without issue albeit only used for a couple of hours each evening but unfortunately the restarts returned. Thermals are fine. High 30s/low 40s for the CPU when browsing. May touch 70 when gaming. 3080FE sits usually between 70-75 when being fully used in game.

I don't have any option for testing a different motherboard or CPU unless I bought a replacement and I wouldn't know which one of these parts I am supposed to RMA as I've got no idea which would be responsible.

What would your next steps be?
 
Associate
OP
Joined
9 Nov 2003
Posts
429
Location
nr Stoke
Power supply is only just above recommended for that gfx card,can you try a higher powered one?
Could it be windows deciding to instal/restart?

It is but my restarts are happening when the GPU will be pulling next to nothing and my CPU is less power hungry than the one I was running previously so the PSU being at fault is something I'd ruled out. Happy to be proven wrong through someone else's similar experience though. I don't have any other PSU to try and the voltages all look good in HWInfo.

Windows definitely not forcing restarts. Slighty misspoken in OP when I said no errors in Event Viewer. There's nothing I can see that points to the cause but I just see 'The previous system shutdown at 14:53:22 on ‎21/‎08/‎2021 was unexpected' & 'The system has rebooted without cleanly shutting down first' when it boots back up.
 
Soldato
Joined
19 Oct 2002
Posts
7,262
Location
Woolyback Country
Possible to test barebones out of case?
incase of a short (unlikely though)
could it be a dodgy power or reset switch? Try disconnecting both and start pc by shorting the power pins on the mobo
 
Soldato
Joined
15 Oct 2019
Posts
11,687
Location
Uk
My money's on the CPU being unstable at high single core boost clocks. Try locking in a all core frequency of 4.4 @1.3v and see it that stops it and if it does then RMA the CPU direct to AMD.
 
Associate
OP
Joined
9 Nov 2003
Posts
429
Location
nr Stoke
Possible to test barebones out of case?
incase of a short (unlikely though)
could it be a dodgy power or reset switch? Try disconnecting both and start pc by shorting the power pins on the mobo

Thought about it. Would certainly be possible to do although hoping to try other options first. The case and wiring are all pre-existing, just removed old motherboard and plonked new one in with updated CPU so like you said, seems unlikely, but won't rule it out.

My money's on the CPU being unstable at high single core boost clocks. Try locking in a all core frequency of 4.4 @1.3v and see it that stops it and if it does then RMA the CPU direct to AMD.

I've read about others doing that so will give it a try as the fact it does it while not pushing the system does point to that.

Just trying disabling hardware acceleration in Opera at the moment as read a few people had restarts with that and it was running at the time of the all the crashes (I think). Unfortunately, I can't force it to crash through normal operation as I've currently been on the PC for over 3 hours, just browsing and it has been fine but it does tend to okay for the first few hours which is why I think I've fixed it in the week when I only get a couple of hours after work but have issues at the weekend when it is used for longer periods. Is there a reliable utility that I could use to push single core performance to try and force the issue?

I've ordered a cheap B450 MB to try if I have no joy with the above.
 
Soldato
Joined
15 Oct 2019
Posts
11,687
Location
Uk
Is there a reliable utility that I could use to push single core performance to try and force the issue?
Try OCCT using small data set SSE instructions 1T, you will have to assign core affinity and test every thread though for maybe 10 mins.
 
Back
Top Bottom