Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Apologies for the length of this question...

I built a new system some months ago, retaining only the GPU and power supply from the old system and have been having intermittent issues with it.

My problem has manifested itself as two issues;

  1. A complete system shutdown - no BSOD, no entry in system logs describing the reason for the shutdown. When I log in, Windows tells me the system was shut down unexpectedly.
  2. Complete system freeze. If audio is playing, whatever sound is still in the buffer is looped repeatedly.
    • If running integrated graphics, I experience graphics corruption when it freezes.

The problem began as a random shutdown, which I attributed to the power supply not being capable of sourcing enough current. I upgraded to a 1kW Seasonic and it seemed to be fixed.

Lately it has been playing up again. It started as random freezing whilst playing games, usually about an hour or so in. Temperatures seemed to be the problem so I used a fan controller utility to force GPU fan speed to 100% and I kept an eye on the temperatures. They never exceeded 60C. This seemed to help reduce the frequency of the problem but I was still having issues with it.

I removed the GPU and tried running on internal graphics for a few days. The issue was still prevalent and the crashes / freezing are becoming more frequent. The system seems stable within the BIOS / POST environment but it has frozen a few times whilst trying to start Windows.

I have updated my BIOS to rev 14 (was rev 7) and loaded optimised defaults, with no success. During the BIOS flash process I was running integrated graphics and experienced freezing with graphics corruption. This also happened when recovering the main BIOS from the backup. A second attempt at recovery and flash proved successful.

System Specs:

  • Intel Core i7 3770
  • 16GB Kingston HyperX DDR3-2133
  • Gigabyte Z77X-UD5H
  • eVGA GTX570 1280MB
  • OCZ Vertex 4 512GB
  • Coolermaster 1000W Silent Pro PSU
  • Windows 7

The system is not overclocked.

There are reports of similar issues from others, however their solutions do not work for me.

In summary I have tried;

  • Replacing power supply
  • Monitoring temperatures
  • Removing GPU and running on internal graphics
  • setting bclk manually
  • disabling C3/C6
  • Flashing latest BIOS

Prior to all these issues I had confirmation that the system was stable from 2 days of Memtest + 2 days of Prime95 without errors. Now I can't even run a memtest in fail-safe mode without the system randomly shutting down.

I'm at my wit's end here. I would like to gather some more information before deciding to ship anything off for RMA and am looking for suggestions as to what to try next.

The system is running on a UPS and it's possible that it is drawing more than the UPS can handle, but the UPS has not indicated an overload condition.


Update 2012-03-31, 18:27 GMT+10

The results of memtest on the individual memory modules is complete.

  • DIMM 0, slot 0 - 9.5 hours walltime, 16 successful passes.
  • DIMM 1, slot 0 - 11 hours walltime, 19 successful passes.
  • DIMM 2, slot 0 - 7 hours walltime, 12 successful passes.
  • DIMM 3, slot 0 - random crashes during POST / memtest.
  • DIMM 3, slot 3 - random crashes during POST / memtest.

I will RMA the memory and continue testing but unfortunately I will have to wait another day due to the Easter long weekend.

My question now is, how can a system be stable, pass 24 hours of memtest, and then suddenly have one DIMM of memory die so unexpectedly?


Update 2012-04-02, 13:54 GMT+10

I have returned the memory to the store for the manufacturer to diagnose, but it will take 3 weeks for them to replace them. Unfortunately I do not have any other DDR3 memory on-hand for use in the mean-time.

share|improve this question
    
What happens if you swap the order of the memory modules (or run it with the bare minimum)? Any other pieces of hardware attached to the system? Could it be the CPU? –  Breakthrough Mar 29 '13 at 15:23
    
You seem to be convinced its a hardware issue. If the system is locking up, maybe it's driver related. Check the Windows Event Logs for anything related to the right before the issue. You could also remove all peripherals and run as basic as possible. Once it's stable, then start adding peripherals until one causes the error. –  CharlieRB Mar 29 '13 at 15:52
1  
@CharlieRB, the issue occurs regardless of the current application (facebook, youtube, gaming, etc) and occurs outside of the OS as well (memtest + QFlash utility) –  bhillam Mar 29 '13 at 16:13
    
@Breakthrough, I don't know why it didn't occur to me earlier to try swapping some of the memory modules around. I'm currently running Prime95 for 50 minutes on a single module (4gb). So far it has passed one round of tests. I will leave it running overnight and see what happens. –  bhillam Mar 29 '13 at 16:15
    
It's probably worth checking your ram by running memtest86 that should find any faults and tell you which stick it's on. –  Col Mar 29 '13 at 17:35

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.