Computer randomly shutting down in the middle of some CPU and GPU activity (viewing videos, gaming)

Keirnoth · Aug 28, 2015

Asking on behalf of someone I know.

His current build -

Processor: Core i7-4770k
RAM: GSkill Sniper Series DDR3-1600 8 GB total RAM x 2
GPU: EVGA GTX 780 SC
SSD: Samsung 120GB 840 Pro
PSU: Corsair HX750
HDDs: Western Digital Caviar Black 1TB

So here's what happens. Usually in the middle of something that sometimes stresses the processor and sometimes doesn't, his display will go blank and the computer will do a hard restart. No BSODs, nothing. He can be watching a Youtube video, or gaming or whatever, and the computer screen will simply go back and restart. He's not even playing a game that stresses his GPU, just Phantasy Star Online 2 at 720p windowed. Nothing big and hardly makes the 780 break a sweat.

Stuff he's already done to no avail:

* Checked bluescreenview for dumps - nothing there.
* RMA the GPU and PSU.
* Run sfc /scannow to look for corrupt files
* Downgrade drivers to older known working versions
* Swap out DVI/DisplayPort cables
* Ran temp monitoring software while gaming - temps come out looking normal for both the GPU and PSU
* Clean reinstall drivers to older known working versions
* Replace thermal paste with new on CPU
* Ran memtest for 4 hours with no errors
* Swapped out positions of memory to the other pair of memory banks
* Plugged power strip into another wall socket
* Plugged computer directly into wall socket
* Updated his bios to the latest release

At this point he only has the CPU, mobo, and memory left to RMA if he needs to RMA, but a brief memtest indicates that it's not the memory at fault.

What else do you think he should do? We're at a loss. My friend was thinking of reinstalling Windows but I'm get a feeling this is a hardware problem that a clean install of Windows isn't going to fix.

The only peripherals he has connected to his machine is his PS3 controller, speakers, keyboard, mouse, and monitor. Nothing else.

What do you think? Should he RMA something or is there something obvious we're missing?

cub_fanatic · Aug 28, 2015

If it wasn't the PSU, it is possible it is the motherboard. This does sound like a power issue and the main lines of the PSU all go through the motherboard so it could be a problem with the power delivery on the board. One thing he didn't try is check for a BIOS update for the motherboard. It is a longshot but if there is one it is worth a try before going through the RMA process.

What case is he using? Has he ever tried running the bare minimum components outside of the case? It could be something weird like the power button or a bad fan triggering a PSU shut down. I'd check any wires coming from the case plugged into the motherboard or PSU like a molex fan connector for example. Also make sure nothing from the case or anything else with bare metal or even painted metal is touching any type of PCB like if you forgot to install the standoffs that separate the motherboard from the case or if a tiny screw is rolling around in there.

Finally, what type of CPU cooler? Is the GPU completely stock from factory or is there some kind of aftermarket cooler like a waterblock?

Keirnoth · Aug 28, 2015

cub_fanatic :

Said in my list that he did a BIOS update.

This isn't a new build, he's been having it for almost 2 years without problems, it's just started happening last month.

The GPU is completely stock, no aftermarket. His CPU cooler is a CoolerMaster Hyper 212 Evo. I'll pass this info along to him, and I'll let him know to run a longer memtest.

BluPulse · Sep 1, 2015

You said he downgraded drivers to know working ones. If its the GPU, the driver should most recent driver. Use DriverEasy to check for updates, and especially look for drivers for connectors and connections on mobo. Something with PCI at start or Standard Bridge are important and should be updated. Try unpluggings fans. Might be disfunctional. Unplug the GPU and use onboard video if there is onboard video through the mobo. Also, if it just started happening, maybe he moved his computer and something got loose? Check all coords going to mobo and hard drivers/ssd. Make sure the 24pin connector is firmly inplace.

cub_fanatic · Sep 4, 2015

Keirnoth :

Not sure if you guys figured this out yet? If it was me and it is still re-starting itself with the motherboard and bare minimum components outside of the case, I'd start with a clean install of Windows before RMA'ing anything. If you have a spare hard drive, you don't even have to do it on the current install which would save a ton of time if Windows isn't the problem. If a fresh Windows install on a different drive fixes it, it could be that the original install was corrupted in a weird way or that the SSD was bad so further examination of the hard drives would be the next step to know which.

When you say he re-did all the drivers, does that also include the firmware for the Samsung 840 Pro? That could be checked via the Samsung Magician app that comes with the drive. It can also be DL'd from their site. Another good diagnostic tool for both the hard drives is HD Sentinel. It will tell you if a drive is failing by showing you all the SMART data.

If all of that fails, the first thing that I would RMA would be the motherboard. Intel CPUs are very reliable in my experience and, typically, it is everything else around it that will fail long before the CPU. If it was the CPU, it would most likely just not work at all nor would it just suddenly start to happen - it would have been a dud right out of the box. Motherboards on the other hand are probably the 3rd most vulnerable part of a high end PC behind the PSU and slightly behind the GPU. As for the RAM, if there weren't any errors in memtest86+ then it is safe to say that is not the problem.

Keirnoth · Sep 10, 2015

cub_fanatic :

We don't have a spare hard drive to test this on, but he's contemplating a fresh Windows 7 install, but we'll do that as a last resort before RMAing the motherboard.

Also I was going to suggest him taking the components out of the case and running it bare, but these shutdowns only happen when he's doing light gaming and/or Youtube/Twitch video watching.

When you say he re-did all the drivers, does that also include the firmware for the Samsung 840 Pro? That could be checked via the Samsung Magician app that comes with the drive. It can also be DL'd from their site. Another good diagnostic tool for both the hard drives is HD Sentinel. It will tell you if a drive is failing by showing you all the SMART data.

I don't think he's having any hard drive problems, is the firmware upgrade procedure for the Samsung 840 Pro painless? I remember trying to upgrade the firmware on my OCZ Vertex and I had to boot from a startup CD to do it.

If all of that fails, the first thing that I would RMA would be the motherboard. Intel CPUs are very reliable in my experience and, typically, it is everything else around it that will fail long before the CPU. If it was the CPU, it would most likely just not work at all nor would it just suddenly start to happen - it would have been a dud right out of the box. Motherboards on the other hand are probably the 3rd most vulnerable part of a high end PC behind the PSU and slightly behind the GPU. As for the RAM, if there weren't any errors in memtest86+ then it is safe to say that is not the problem.

We're currently swapping out memory sticks. We already tried putting the memory sticks in the other two slots - no go. He's currently running with 1 of the two sticks and "testing" it that way (he really does not want to leave the PC on for 24 hours to memtest it, he doesn't have an alternate PC to do his stuff on). If he can run with one stick for several days then we can conclude his memory's the issue. At the same time, we're having two sticks coming in (different brand) to replace those. GSkill unfortunately doesn't do advance RMAs.

We'll keep you all posted. Thank you all for the help.

cub_fanatic · Sep 10, 2015

FW updates for Samsung SSDs are very painless - just load Magician up and click on the firmware update tab.

The fact that you say it is happening only when gaming or streaming video, is the gaming online as well? Did you try to use a different wifi adapter or ethernet card?

Keirnoth · Sep 15, 2015

cub_fanatic :

After bashing our heads against the wall several times, we concluded it was bad RAM sticks. Got new RAM to replace the old faulty ones (old ones were GSkill Sniper DDR3-1866) with new Corsair Vengeance Low Profile DDR3-1600 RAM and now it works like a charm.

This is the first time I've ever seen bad RAM cause a computer to *just shut down*. All signs pointed to the PSU but I guess the darnest things happen, eh?

cub_fanatic · Sep 15, 2015

Keirnoth :

That is odd that they wouldn't show any errors during memtest. In your first post, though, you said it was 1600 RAM but it was actually 1866? While most mid to high end 1150 motherboards should support 1866 and higher (technically as overclocked speeds), the 4770k actually only officially supports 1333 and 1600 according to Intel. So, it is possible that the motherboard you were using the 1866 in simply didn't like to overclock its RAM.

Search

Computer randomly shutting down in the middle of some CPU and GPU activity (viewing videos, gaming)

Keirnoth

Honorable

cub_fanatic

cub_fanatic

Honorable

Keirnoth

Honorable

BluPulse

Reputable

cub_fanatic

Honorable

Keirnoth

Honorable

cub_fanatic

Honorable

Keirnoth

Honorable

cub_fanatic

Honorable

TRENDING THREADS

Latest posts

Moderators online

Share this page