Random Reboots - Not on windows, only Ubuntu

[samuel.piecz]

Hello everyone.

Specs:

CPU - i5-4670k
GPU - GTX 1080
PSU - 750W - Don't remember brand, it's not a shitty PSU though.
Motherboard - msi z87 g45
Ram - 8gb DDR3

I have dual boot windows 10 and ubuntu 16.04. I recently upgraded my GPU to a 1080 and overclocked my i5-4670k to help out with it when triple monitor gaming.

Prior to my upgrades, I had a single gtx 760 GPU. When working with linux, or windows, I had no reboots and everything worked great. I picked up a second gtx 760 and ran it in SLI. I have a 650W PSU and experienced random reboots in Ubuntu, but not windows 10 for some reason. My assumption was that Ubuntu's drivers were simply not as efficient and that was the reason for my reboots. To fix the problem, when I was not gaming I would just take on 760 gtx out and Ubuntu was fine.

I was pretty tired of removing and putting back in my graphics card, and could use an upgrade anyways, so I picked up the gtx 1080. It works great in windows, and great on ubuntu. However, I experienced some stuttering on certain games, learned about overclocking, and picked up a non stock air cooler for my CPU. Noctua for those who are wondering.

My cpu temps are fine and stable, I ran IBT 5 times on high settings and had no BSOD. Windows is still working fine. Ubuntu seems to have an issue though, the CPU temps never exceed 56 degrees, so I don't think it's overheating. I'm wondering if it could be due to my overclocking settings? I did some digging in Ubuntu forums and checked my logs, and I don't have any errors. So I came to the conclusion it must be a hardware issue.

Posting this now in case my computer restarts, don't want to lose all that I've written. Going to reboot PC and take pictures of BIOS so you can see if my OC settings could use updating.

Thanks a bunch in advance for any advice / suggestions.

Edit:

Here is a link to images of my BIOS settings for my CPU
https://imgur.com/a/0vv9IbT

Sam

 

[mdd1963]

The overclocking, if slightly above/past/beyond the ragged edge, can induce corruption on whatever OS was running at the time of the instability, and, remain there randomly causing crashes until the offending corruption is deleted/quick formatted/OS reinstalled...

Or, your PSU, regardless of rated wattage and name brand, can becoem marginal over time, just intermittent enough to make it hard to blame...

Likewise, a very intermittent RAM module can be difficult to narrow down, easiest to simply swap with known good RAM for testing....; there are no certainties that any intermittent tendencies of RAM will materialize at the 1-2 hour periods you happen to be running stress tests, RAM tests, MEM test, etc...

Last, but not least, some motherboards can develop marginal voltage regulation, and, work with some GPUs, but, not with others that might have much more modern but allegedly less wattage GPUs, as the power delivery changes for some newer GPUs might be MUCH more rapid than some motherboards can handle... (seems hard to fathom, a mainboard handling a pair of older 150 watt GPUs fine, but having issues w/ a single 130 watt, etc..

Summary--at stock timings/clocks, delete partitions, fresh install, retest for stability. (if you have a new SATA drive and cable, try it, just to rule out as many potential culprits at once) Try/borrow new known good RAM at semi-conservative timings (just becuase 3666 MHz worked last week does not mean it works today), and /or a known good PSU, test GPU in another rig with extended playing...