7days crashing quite often. Please help.

tenchikun said:
Not sure what else I can do unless someone has any other ideas
Click to expand...
Since you did a clean install of windows, I kinda suspect a hardware problem.

Here's a few things that you haven't mentioned checking:

1: NVME temperatures (these can overheat and cause a crash) The latest version of Windows 10 has built in a way to check temps. Check it while playing of course.

     It's happened that system builders have installed an nvme drive with a heat sink, but forgot to remove the plastic layer from the sink, which makes the drive even hotter.   So if it is running hot, I'd take it out and check the heat sink is properly mounted. (I would check it even if it's not running hot)

2: Test and repair all your disks with CHKDSK /F or /R

3: Check power distribution to your mb (xtra 4/6 or 8 pin plug/cable near the cpu) Connecting this cable has sometimes been overlooked by a builder. System still works until it's stressed.

4: Power distribution to your GPU (make sure 6 or 8 pin cable is securly connected)

I would also try running furmark and prime95 at the same time to see if I could get the system to crash. If successful, this would clearly show that it is a hardware error.

Leave a few cores free in prime95 when running both tests simultaneously, so the GPU has something to work with.

Good luck.

 
So I ran Furmark and it runs fine on 1080p but if I set the resolution higher than that it freezes. Doesn't stop responding in the task manager at all. But it's obviously crashed and the event view gives me a warning nvlddmkm stopped responding.



 

ssd temps are fine. tested the disks, no errors, 8pin 12v connect is plugged in and secure, same with the gpu. 

 
tenchikun said:
https://i.imgur.com/aYd0CiW.png I found it odd that raytracing is enabled on a non RTX card. 
Click to expand...
I get the same on my 1060. It was made available via the drivers a while ago. If a program doesn't use it, it won't have any effect.  If a program does try though, be ready for some FPS issues. I tried the Quake RTX demo. Max FPS was 5, with the average being below 1.

Our systems aren't that far off from each other, so I'm at a loss to understand some of the issues you're having.

 
sorry if I missed it, but:

Did it work correctly some time before, and then the crashes started to happen,

or did it occur right from the beginning when you first tried to play the game?

Did you recently change/upgrade hardware components?

Please give the specs of your PSU (or vendor+model) and how old is it?

 
Last edited by a moderator:
tenchikun said:
So I ran Furmark and it runs fine on 1080p but if I set the resolution higher than that it freezes. Doesn't stop responding in the task manager at all. But it's obviously crashed and the event view gives me a warning nvlddmkm stopped responding.



 



ssd temps are fine. tested the disks, no errors, 8pin 12v connect is plugged in and secure, same with the gpu. 
Click to expand...
Sounds like either GPU or PSU issue to me, and I think it's hardware related, not software. If you happen to have a spare part lying around, maybe try swapping one of them out and see if the issue happens again. Though, keep in mind that swapping out a GPU with one that uses less power might not help much since if it is in fact a PSU issue, using less power could help prevent a crash.

 
I was going to suggest a full DDU driver uninstall, but since you already wiped and reinstalled Windows that's a dumb suggestion. Still have your old 1050 lying around? Maybe you could also try the 1660 in the other PCI slot. Throwing darts at this point...

 
tenchikun said:
power supply is an evga gold 650watt psu
Click to expand...


tenchikun said:
8pin 12v connect is plugged in and secure, same with the gpu. 
Click to expand...


I think your power supply also has connections at the psu. So you could check the connections there as well.

tenchikun said:
So I ran Furmark and it runs fine on 1080p but if I set the resolution higher than that it freezes.
Click to expand...


This seems to point to a video card problem so maybe replace that power cable from the psu to the gpu. (if you have an extra one)

Or try the cable in a different output from the psu.

And pull out and reseat the 1660.

 
I've already pulled the card, reseated it, unplugged the power cable and replugged it in. Someone else had suggested that to me. it was one of the first things I had done. sadly I don't have a spare cable. 

I don't have the 1050 anymore. used it in my dads computer when I upgraded. 

 
Just shut down. unplugged the video card and reseated it and connected it again. Figured why not try it again. Still no dice! 

 
https://i.imgur.com/thJ35Bo.png

I photoshopped this log file together so you could see what I noticed during the Furmark test when it locked up like 7days does. At the point of the crash the gpu frequency spiked to 1935mhz and the gpu voltage went .85 to 1.03. The gpu shouldn't go over 1850 boosted.

 
So on a good note, I underclocked my gpu with msi afterburner. -280mhz and ran furmark with 8x antialiasing and 1440p. Ran for like a half hour and no crashes. So I kinda think this might be a driver issue causing the gpu to suddenly overclock somehow and become unstable. I don't know how else that can happen. This card used to be stable and then it seems after some driver update, that's when all these problems occured cause it happens with outriders, call of duty black ops, valheim, and red dead 2. 

 
I realize your PSU is high quality, but how old is it? Tech channels seem to recommend replacing them every 5 or so years depending how much you use them. Capacitors wear down over time, and as they wear down, the max rated wattage output also goes down + or - 10% per year depending how much you use the PSU. Of course, gold rated PSUs tend to have the solid caps, so I imagine the percentage is lower than 10%. The other thing is that 650 watts is not exactly a lot of leeway given the new power requirements of the upgraded hardware. Sure, it's more than enough in theory, but over time, as the caps degrade and efficiency goes down... I don't know.

Also, underclocking the GPU may have lowered the overall power usage a bit which may have helped a little which is why you thought you figured it out at the time until it crashed again.

 
Last edited by a moderator:
tenchikun said:
So on a good note, I underclocked my gpu with msi afterburner. -280mhz
Click to expand...
Try underclocking the vram. Also maybe try undervolting slightly. (both in Afterburner)

Also try limiting the frame rate to something ridiculous like 45-50 fps using the Nvidia Control Panel. (just as a test)

 
tenchikun said:
The only thing i've upgraded was the mobo from an asus rog b350f to a b550f-wifi. And from a gtx 1050ti to a 1660ti. Had no issues with the previous set up. the power supply is an evga gold 650watt psu.  Oh and I upgraded from a ryzen 5 to a ryzen 7. 

https://pcpartpicker.com/product/y88H99/evga-supernova-g3-650w-80-gold-certified-fully-modular-atx-power-supply-220-g3-0650
Click to expand...
So you "only" changed your whole platform (CPU, GPU, mobo).

Sorry I have to ask more specifically:

You said "fresh new install of Win", meaning what?

In your case it should have been a total new install to a fresh empty partition on your harddrive. Everything else (repair install or via some "migration tools") might cause all kinds of undefined bs-behavior of your system. No cleanup tool or similar can then provide a fully functional Win on that drastically changed hardware (esp new mobo).

Insufficient power supply can cause same, but your PSU should be more than enough,

even without further details given about your CPU..."I have a Ryzen 7" is ridiculously (no offense meant) unspecific...next to possibly other peripheral devices built in your rig.

...at least you could confirm that you haven't running a handful of USB water boilers 😄 just kidding, PSU can be dropped from the list I guess. 👍

 
Last edited by a moderator:
You must log in or register to reply here.
Back
Top