PC crashing randomly, BSODs, hot GPUs, I got it all.

Hardware and technical stuff
1

I just want to say that I hate PC gaming sometimes.

My Rig, which I bought at CyberPowerPC, because I was too scared to build my own in 2013 is dying yet again. With pretty much the same random issues I saw in May of last year. Computer freezes up, and crashes. Playing a game, gray screen with audio for a second, restarts.

Awesome.

Now, because the Powercolor RX 280X that I got as an RMA replacement already died earlier this year, I got a used GTX 670 to last me until the new summer GPUs dropped. While I was playing WoW, I noticed that the GPU temps were getting CRAZY hot, like 97C, and, of course, the fans were not at full speed, for some reason. I installed the overclocker software, and set the fans to 80%, and the max load temps for the GPU was 85C, hot… but not scary hot. Old reference GPU, it runs hot, of course. After turning off my PC, I opened her up and everything was super hot, for a watercooled CPU, it seemed odd. I am not overclocking (the water cooling upgrade was free), I was running HWmonitor, and the other CPU and mobo temps were not even hitting past 55C

So, I assumed this was a heating issue, the small stove I had burning in the upper PCI-e slot was heating up my mobo and CPU, causing the crashes, or just crashing itself. So, I moved the card to the second PCI-e slot, and figured, with the extra room in the case, and the buffed up hairdryer fan speed, I can hold out until an RX 480 (the 4GB one) could come in the mail.

I thought nothing of it, and went to sleep. Headed to the PC to play some more WoW:Legion on this long weekend, and noticed it had shut down. Ah! I had left handbrake on overnight to rip my simpsons DVDs, I must have checked the “shutdown” option!

Nope.

I checked the event log, and sure enough, there had been kernel power events every 2 hours overnight! Computer was restarting itself, and then crashing before I even logged in… all night.

So… what the fuck is going on with my computer? Is this a hardware issue? I hesitated reinstalling windows, because that would be a huge pain, if the software wasn’t the problem.

Now, back in 2015, the issue I was having was clearly GPU related, screen would go all colors/bars and PC would shut off, and there would be garbled video. This time around, aside from getting a quick gray screen when shutting down in game, when I have been using the PC for browsing, etc, the crashes happen with a BSOD, that goes away so fast I can’t read. So, it seems different.

I am wondering if it is a mobo issue again.

Quick Questions:

  1. How does installing hard drives work for a new computer, like, if I were to take out the motherboard and CPU, (LGA 2011, no replacements around) and get a new i5 6500 and a motherboard it fits in, could I just hotswap everything else out but the mobo and the cpu? Would the windows install be ok with that?

  2. Can you think of any issue it could possibly be? The event viewer just shows “kernel-power” with no helpful information.

  3. Am I allowed to vent on here? That was a bit rambly-posty.

  4. Should I just buy another stusser-beast candidate? I mean, the XPS-s they have now for about 800 bucks seems like nearly what the cost of replacing the mobo and CPU will be (and likely the ram)

  5. Would a new socket mobo take my old ram? Can the LGA 1151 take DDR3? I think some can right? Is it backwards compatible?

One final comment. Dell will forever have my respect for the fact that I am typing this out on my Stusserbeast™ XPS I bought in 2008, that is still kicking after all these years. They build some quality stuff.

TLDR: PC issues again ruin a long weekend of gaming.

2

Venting is the least of your problems! Oh wait, actually it might be the crux of your problems. :)

I will say that unless Windows has become much more forgiving in recent years, you will not be able to install a new motherboard and CPU without reinstalling your OS.

As far as your problems go, it certainly could be heat. I wouldn’t think 97C on the GPU would be enough to cause shutdown, but it’s not ideal. Even if that were an issue it wouldn’t explain your late night reboots unless something was causing the card to throttle up outside of a 3D environment (or the fan isn’t working properly either due to hardware or software malfunction).

I think you should look closely at the power supply. Is its fan working properly? Is it running really hot?

Generally speaking when I hear mystery reboots the first things that come to mind are:

A: Power supply failure.
B: Memory errors.
C: Bad motherboard.

3

I remember Rock8man reporting very high temps with his Cyberpower PC too. I’m noticing a possible trend. Does the case have room for additional intake or exhaust fans?

I haven’t played WoW in a long time, but I didn’t think it could push modern hardware to the limit. Does WoW leave the frame rate uncapped by default? Try enabling vsync if it’s not already, and check your GPU temps again.

Unfortunately, I don’t have a more specific diagnosis for your PC than “it sounds heat-related,” which you probably guessed already.

As for the hard drives, they’re easy to swap. SATA connectors are standardized, just plug them into the new motherboard. Not sure about the RAM, but the new stuff is really inexpensive these days.

4

Yeah, high temps…when I was told in the marketing that they did excellent cabling to improve airflow. I don’t think the heat related issues would manifest during overnight reboot sessions. The GTX 670 is shitty and slow, but I have been using it for months without issue.

Either way, WoW:Legion uses a bit more juice, and I do believe the framerate is uncapped.

The PC works for an hour or so before BSODing.

The PSU is a good idea to check. As I have no spare parts to muck around with, I will have to take this in to a shop to have them part swap and check components.

5

Oh, one other thing. Turn off the BSOD automatic reboot, it may help with the diagnosis.

6

I have nothing to add except that as someone suffering from mysterious PC overheating, I feel for ya, man. My machine started overheating a few days ago and it’s so old that I decided piecemeal diagnosing and fixing wasn’t the best use of time or money. So I ended up buying a rig I’d been looking at for a while:

https://www.amazon.com/CYBERPOWERPC-Xtreme-GXiVR8020A-Gaming-Desktop/dp/B01HNBLHAA/ref=sr_1_1?ie=UTF8&qid=1473158048&sr=8-1&keywords=cyberpowerpc+480

Now that you guys are all relating how your CyberpowerPC rigs are overheating… that makes me a teeny bit concerned. But aside from that, it seems like a relatively good machine for the price. Plus, it’s got a freakin’ RX 480 in it… will be quite a step up for me.

I’m planning on cloning the 1TB drive to my existing SSD and making that the boot device. Maybe I’ll upgrade the RAM… it’s not in the short-term plans though.

I hope either the sympathy or product recommendation helps!

7

First port of call would be to verify which part(s) are overheating and in what circumstances. Use different monitoring software to make sure it’s not misreporting a sensor. Check within BIOS if necessary to see if it still happens when completely idle. My instinct says it’s not a motherboard problem. I’d rule out dust, faulty/dry watercooling, dodgy fans, crapped out paste and iffy PSU long before I worried about the motherboard with these symptoms.

Cabling is not going to make the difference between normal and catastrophic temps if you’re not overclocking, or even realistically if you are. It might get you a couple of degrees.

8

Well, I am of course Anecdata. My brother, who after finishing pharmacy school, bought himself a present of a cyberpowerpc computer, and his is humming along fine, replaced the video card with a 970, added an extra hard drive. And he has had his PC for 4 years.

Cyberpowerpc tip - Don’t buy the warranty, they will charge you to ship out your PC, many horror stories of paying hundreds to ship out the PC, only to have them send it back broken.

I have checked in the bios, and temps seem normal there, but I haven’t left it on for extended periods.

Also, would turning of BSOD automatic reboot even be worthwhile? The glimpses of the BSOD screen I got didn’t seem helpful, and isn’t that the same data that shows up in event viewer?

I don’t know how to tell if my watercooler is dry. It says if the light on top of it is on, that it is good. In the manual.

Anyway, I think I am just going to take the thing in to a local shop to have them do the diagnosin’.

9

Maybe that’s the best advice?

10

I was having some BSOD issues for a while and found a helpful program called BlueScreenView. It will analyze your BSOD memory dumps and give you info that may help you track down the problem.

11

Bluescreenview doesn’t show anything for me. I don’t think it is getting crash dump data from the right spot.

Anyway, I turned off automatic restart, and I was greeted with a teal? screen of death that said.

error: KMODE_EXCEPTION_NOT_HANDLED (storport.sys)

Whatever that means.

My google-fu brought up, it could be anything, but that is dealing with storage drivers or something, a driver crashed. Could be hardware or not? Anyone have any ideas?

12

I was going to mention WhoCrashed earlier but I figured it was a hardware issue. If it’s possibly driver related then definitely give that program a look.

13

The gist of what I have been reading, this is due to Intel Rapid Storage Utility, which I do have installed, (carried over from my original windows 8 install) and drivers crashing. Uninstalling that might help.

Interestingly enough, I ran handbrake last night, to continue the conversion of my DVDs into mkv files, and my CPU load temps got up to 75C, but that was with all cores at 99% power, but it was only 5 minutes of ripping. I am also wondering if the watercooler I got isn’t working anymore. That is kind of hot for an i7-3820, but not dangerously so.

Intel says, Tcase is 66.8C

Hmm, the RPM on my fans didn’t go up during the encode, stayed at 1300, I wonder if my CPU cooler isn’t working very well either, but encoding via handbrake is 100% CPU load. But, in 5 minutes to shoot up to 75C seems… bad.

Will attempt to uninstall the Intel Rapid Storage drivers (which do not list windows 10 as supported)

14

Removing the Rapid Storage drivers sounds like a plausible place to start. I would also recommend cleaning the inside of the PC thoroughly (use compressed air and remove any components that are removable so that you can get at all the little places dust and debris like to hide) paying careful attention to fans/intakes/heatsinks/etc… I recommend doing that every few months anyway.

Before progressing beyond that and spending any money though, I’d highly recommend downloading Memtest86+ and creating a boot CD or USB with it. Run the full tests against your installed RAM. Oddball sudden unexplained reboots and crashes are a hallmark of a RAM chip going bad. Your description of getting an hour or more of use out of the machine before seeing it crash leads me to suspect memory as well, since the PC could work fine for hours right up until the particular problem range on the bad RAM chip was addressed as the machine paged more RAM for gaming, DVD conversion, whatever…

At the very least a thorough test of RAM is an easy and FREE diagnostic you can do yourself to rule out one of the most common causes of crashes and reboots.

15

Ran windows’ version of Memtest already, all turned out good. I could try using that software.

I am starting to think it could just be driver related, as there were a few windows updates recently.

16

Just make sure you don’t need the RST driver before you uninstall it. It may be necessary if you’re using your motherboard’s RAID controller, for example.

17

Nah, I am not doing RAID things at all. Otherwise, I think i am first going to check ASUS’ website for a driver update, it is very likely I could update it and fix it.

Unless of course my motherboard is again on the fritz, and this is just a symptom of the larger issue.

18

If I were you, first I’d see if popping some fresh new RAM in there fixed things. I had some bad RAM for about 4 years that caused random BSODs and yet always tested fine with memtest86+, but I finally tried Prime95’s “Blend” test and almost IMMEDIATELY got errors, errors which vanished as soon as I got new RAM. I’m still kicking myself for thinking that memtest86+ was the be-all and end-all. It’s not.

Also I’d see what I could do about better cooling in your case, but I think you’ve got “stealthy” bad RAM like I had.

19

So yeah, removed the CPU fan, and the water cooler radiator was caked in dust, cleaned it off with a dry rag, and air dried. CPU temps are lower, still getting crashes though.

I disabled the RSTe at startup, so we will see if that does anything. I don’t need it, no raids here.

20

Try the Prime95 “blend” test (Prime95 is free BTW) and if one or two of the threads (the program calls them “workers”) craps out (give it at least an hour or better two, unless you get a BSOD first of course), consider trying the same test without one of your DIMMS (assuming you have 2), re-running the test, wait till it craps out again–if it doesn’t in two or three hours you’ve found your culprit. If it does, test the other stick. Even if that stick checks out, assuming that the sticks came as a kit, you should send for a warranty replacement since RAM kits usually have a lifetime warranty. On the other hand, if the RAM came as part of a prebuilt system, you’re pretty much boned at this point if the system is out of warranty, and you should just order a new RAM kit from Amazon or Newegg or wherever, install it and re-run the test.