System freezes with no dump or error

[alexsan]

Hello all,

I got a brand new PC build and I am having the issue that it somethimes freezes. I cant use the mouse and keyboard when it happens and there is now BSOD or error message. Also the eventlog show no errors or warnings prior to the time it happen.
I have to physically reboot. I left it to see if ot would respond over some time but it just stauys in the "frozen state".
There is no unussual activity from high spinning fans or disk activity, it looks just idle and temperatures seem oke.
The issue occurs randomly when working in different applications or playing different games.

Every time the issue occurs I here a weird short sound over my headset (2 seconds of an irritating sound) At first I thought It had to do with the onboard audio/drivers so I disabled the audio in the BIOS and plugged in my USB headset, the sounds is still played when the issue occurs.

I hope you guys can help because I really dont know what the cause could

Below is an overview of my system specs. All drivers are updated to the latest available version and I also updated to the latest BIOS version which didnt solve it also.

(One thing Im still looking into is my BIOS setup. I only enabled XMP profile and it caught the correct RAM timings and speed but it looks like it also sets a target speed for the CPU (3.9). Not sure why this is happening)

- Mobo: Asus Maximus VI Hero (latest drivers and BIOS)
- RAM: 2 x 8 GB of G.Skills Trident-X (2133Mhz)
- CPU: 4770K
- CPU Cooler: Noctua NH-D14
- GPU: GTX 680 (MSI Twin Frozr III) (Running latest nvidia driver)
- PSU: Seasonic x-750
- Case: Corsair 750d

I am running an official Windows 8.1 home premium with all updates installed.

What can it be,.. what can it be?

Thanks in advance!

 

[Dr-Kiev]

When screen freezes, the most obvious culprit is videocard

 

[alexsan]

I replaced the video card with the one in my old system, im sure that one is ok but the issue stays.
I also tested the RAM 2 times with memtest86 for 7 cycles. It took a while but it came out clean both times.
I reinstalled Windows over 6 times but stil no luck.

Now ehen I run Prime95 after max 10 minutes of running the system just shuts down, without an error message or BSOD dump.
The eventlog only shows a id 41 Kernel-Power error.

I have no clue what the problem can be.

 

[Ketchup]

I usually only hear the sound you are describing when I would overclock too far.

Are you overclocking your CPU? If the BIOS by default is setting the CPU speed to high, manually set it to 3.5 and see how things work out.

What do your temps look like?

 

[alexsan]

Im not overclocking, BIOS is default and Im only using XMP profile 1 for the RAM to get correct timings and speed.

I just replaced the PSU with an other Seasonic and im sure that one is good. Same issue.
Last thing Im trying now is reinstalling Windows on the WD black HDD, I disconnected the Samsung SSD.
Will run the same test again to see if the SSD is the issue.

If not I guess its either the CPU or the motherboard.

Is there any way I can identify which one would be the issue? I can not send them back both I think.

 

[Ketchup]

Im not overclocking, BIOS is default and Im only using XMP profile 1 for the RAM to get correct timings and speed...

That's fine, but in the first post you mentioned that it

...sets a target speed for the CPU (3.9)

in which case you would be overclocking.

 

[john3850]

Try with only one stick of memory.
This almost reads like some other posts were the socket pins were bent.

 

[ericloewe]

If I understand the problem correctly, I had the exact same problem until a few days ago.

For reference, the problematic setup:
Core i7 930 at stock clocks, cooled by an Arctic Freezer 13 and later a Corsair H100i
Gigabyte X58A-UD3R
Triple-channel Corsair Dominator memory rated for 1600MHz
SLI'd GTX 460 1GB (with no slot between them - this was most likely a mistake)
OS drive is a Samsung 830

At first it would happen intermittently, mostly when not stressing the GPUs. After some time it developed into consistent crashes when stressing them. Sometimes, after the required reset, the BIOS would complain of memory overclock failure (I keep it at 1600, Nehalem officially supports 1066) but several Memtest runs (Must've been ~20 over three days, on three seperate occasions) revealed no issues. CPU temperatures were always acceptable.

The problem was fixed by the GTX 780 that replaced the two 460s.

tl;dr If you can, try removing your GPU and using the iGPU instead - the 680 may be malfunctioning

 

[Bubbleawsome]

PSU? I had random shutdowns and freezes when my PCI power cables were screwing up.

 

[Matt1970]

Sure sounds like power supply.

 

[John Connor]

In post #5 he said he replaced the PSU.

 

[denis280]

Sure sounds like power supply.

Yeap! sound like it