Software Tests to isolate problem, what to do in case of hardware failure?

[Leinahtanwc]

I am running into several problems, all of which are making me uncomfortable at the moment. I built a new computer for someone in my family and its not working as smoothly as I hoped. I mentioned previously that the computer was doing an odd boot-up sequence (http://www.tomshardware.com/answers/id-2878567/odd-boot-fresh-x99-sabertooth.html), and now I have run into 2 Blue Screens of Death. While the boot-up problem can be easily recreated (but has yet to be explained), the BSOD's are more random, the first occurring after installing several programs such as Microsoft Office, Printer, etc., the next time while VPN/Network Drive Mapping. I have been suspecting it is the Motherboard, ever since the first bootup - Bios Crashed when I tried to change the clock, the Bios seemed to freeze/lockup/skip every 10 seconds as if it was refreshed, etc., but I would like to run a couple of benchmark tests to locate the problem.

I am looking for CPU, GPU, Memory Testing software, etc., since I want to rule out everything I can before I "confirm" its the Motherboard or some rogue driver corruption. I would trust people here on Tom's Hardware over the google searches which seem to supply dubious tests from sites I have never heard of.

Also, in the event of faulty hardware, who here has experience in dealing with RMA's (With Asus, Intel, Mushkin, EVGA)? As of now, I do not believe I can return any of the products to the stores I picked them up from due to packaging (all items may not be accounted for, lots of cardboard, packaging thrown out, etc.). All products were purchased via Amazon and Fry's. I could really use some help here since I am looking at 1300$ investment fizzing here...

For those who are curious, the BSOD's are reporting Driver Corrupted Exppool, which could be many things from what I have seen.

 

[BrandonYoung]

I know very little about your specific scenario, but it sounds like a cheap PSU. Running the CPU frequency too high could also cause such issues. If you failed the silicon lottery, perhaps the CPU at stock speeds is not stable (this would be a QA failure).

I would start by posting specific hardware for the build as a good place to start.

Some things you could try in the meantime: disable any overclock, return voltages to stock settings, and perhaps disable any "Turbo Boost" types of settings at least until you can isolate the issue.

 

[Leinahtanwc]

I know very little about your specific scenario, but it sounds like a cheap PSU. Running the CPU frequency too high could also cause such issues. If you failed the silicon lottery, perhaps the CPU at stock speeds is not stable (this would be a QA failure).

I would start by posting specific hardware for the build as a good place to start.

Some things you could try in the meantime: disable any overclock, return voltages to stock settings, and perhaps disable any "Turbo Boost" types of settings at least until you can isolate the issue.

CPU: i7 5820k (Clocked up about 18% via quick tuner)
MB: Asus x99 Sabertooth, was on Bios 1801, now updated to 2001
PSU: EVGA Platinum 850W P2 SuperNova
HDD Primary (Boot): Samsung 950 M.2 Pro 256 GB
Mem: Mushkin Redline Enhanced DDR4 2800 (recommended on Tom's Hardware)
GFX: 560 GTX ti (Used, but was working before)
Case: Corsair 760t Black

I didn't go cheap anywhere since I didn't want anything to fail. For now, I would like recommendations for running tests (stress tests, memory tests, etc.) so I can rule out problems with those. I would also like to find a reliable registry fixer/cleaner in case some of the software installed ruined a driver in some way.

As for Software, I installed BitDefender 2015 (2016?), Adobe Acrobat Pro X, Microsoft Office 2010, etc.. Some of the software installed is not new since this computer is replacing an old one - so there does exist the potential that an old program is tinkering with Windows 10 Drivers in some way.

I also installed CoreTemp. I did notice that it didn't display the CPU model like it does on my current computer, but it does list the CPU temps for the 6 cores.

Everything is on Auto except for the CPU overclock, which is at about 16-18% (don't have the exact value memorized)

 

[BrandonYoung]

To get back to your initial question then, MemTest86 for testing RAM, Prime95 for testing CPU (with some RAM as well).

If you care to test your GPU, you could try FurMark to really put stress to it.

For your SSD, Samsung Magician is probably the best choice.

If you would like to rule out software issues such as drivers or a mucked OS install, perhaps boot some flavor of Linux via USB and see if it works correctly.

I am glad you didn't go cheap on your PSU, as it is (in my opinion) the most vital part of any computer build.

I strongly advise against "auto overclock" features such as 'quick tuner' and others, as they typically do a very poor job at overclocking, there are many guides that can help you get a stable overclock (assuming your components are able to be ran stable initially).

 

[Leinahtanwc]

Thank you for the programs. I will download MemTest86 and Prime95.

As for the PSU, I got it on sale for 140$ during the Black Friday on Amazon. Usually that thing is 180$+ for the 850 Model. I was originally aiming for the 750 Model but since the 850 was cheaper due to the sale, I took it instead. I wanted either the 850 or 750 since the CPU socket on the Motherboard used 4 + 8 Pin. The 650 only had one 8 pin.

I will in the near future manually set the overclock settings. I noticed the XMP settings were actually lowing the clock to below base values (dropped the 3.30 Ghz to 2.80).

Now for Driver and Registery cleaning. It is odd to have to do this in so close a fresh OS install, but maybe I should get a program to check driver/registeries. Would CCleaner do this? CCleaner with another program? CCleaner has been helpful on my own personal computer but I don't know if it would be helpful here (I am trying to use only a few programs at the moment to reduce potential problems).

 

[BrandonYoung]

I'm not sure about CCleaner personally, if I have instability due to software issues that can't be resolved by reinstalling drivers, I typically format and reinstall the OS.

 

[Mark_1970]

Can you post up the C:windows/dump file here


If you have recently installed any new software, check to see if it is properly installed. Check for updated drivers on the manufacturer's website.
To debug this error, use the special pool option of Driver Verifier. If this fails to reveal the driver that caused the error, use the Global Flags utility to enable the special pool by pool tag.

For information about the special pool, consult the Driver Verifier section of the Windows Driver Kit.

 

[Leinahtanwc]

Well, the computer now suddenly rebooted with no warning it seems. We were not at the computer when it rebooted but we did arrive back in time to see it on the boot up screen going into Windows 10. Some files were lost, and MS Word reported it tried to save files due to power loss.

There is definitely something wrong now, hardware wise. I suspect it is either the Motherboard or the PSU at this point. I called up EVGA and they recommended I use Nirsoft's Blue Screen Viewer to look into the Dump Files. Anyone have any experience with NirSoft? I also registered my power supply with EVGA in preparation for an RMA. I am also going to call up Asus tomorrow about the problems I have been having.

To let everyone know, I did run MemTest 86x on the memory, one time via Bios boot, one time via UEFI, and both times it reported no errors in the memory. So it should be safe to say that the Memory is not a problem. GPU has not been tested yet, but it was my old GPU for years and I do not suspect it to be the problem. Any input from anyone else?

 

[Mark_1970]

Can you post up the C:windows/dump file here

 

[Leinahtanwc]

When I can, I will. Right now the computer has to be used, despite being faulty.

*Wait, you want the big multiple gig file or just the text log? I don't know if I can upload the whole dump (everything that was in the memory at the time of the crash).

 

[Mark_1970]

When I can, I will. Right now the computer has to be used, despite being faulty.

*Wait, you want the big multiple gig file or just the text log? I don't know if I can upload the whole dump (everything that was in the memory at the time of the crash).

Text log, it's tiny, Actually don't worry about it

 

[Leinahtanwc]

I now have access to the computer in question. It had another BSOD, this time we saw it (and I had disabled the auto-restart when such an error occurs). I also downloaded the NirSoft Blue Screen Viewer program, which has revealed that there have been a total of 6 BSODs so far. Timestamps on the BSOD's show that the program does have it right and confirms the earlier sudden reboot was a BSOD as well.

5 of the 6 BSOD's are reported as DRIVER_CORRUPTED_EXPOOL. The other Bug Check String was REFERENCE_BY_POINTER (this was the first recorded BSOD).

All 6 BSOD reports seem to point to ntoskrnl.exe as the culprit (highlighted as pink in Nir). Crash address appears to be ntoskrnl.exe+142760.

Other files that appear in pink: acsock64.sys, CEA.sys, and FLTMGR.SYS. It should be noted that these 3 files only showed up once each in pink, leaving 3 of the reports solely with the ntoskrnl.exe as the potential culprit.

If anyone has any idea on where to go, I would greatly appreciate it. I need help with figuring out what to do. I have these dump files but I don't know how to upload them/txt post them.

Am I looking at a hardware problem or a software problem at this point?

 

[Leinahtanwc]

Two more BSOD's. They are becoming more regular now and are now occuring after just an hour of use or less.

Here is more information: After googling the driver/kernal in question, people stated that overclocking was the culprit, so I went to the BIOS and reset everything to default. However, the computer locks up shortly after windows loads when I do so. When I overclock it (via EZ tuner), just saying the computer is for work with a tower cooler), the computer is able to load Windows Normally. I described this scenario to the Asus Tech support and he suspects the memory is not getting enough voltage and that the issue may require a bios configuration.

This is driving me nuts!

*Someone has hypothesized that maybe the M.2 Hard Drive is the culprit... Is that even possible? Reasoning is that the drivers are stored on the M.2 and that it may not be accessing them properly from there.

 

[Leinahtanwc]

Well, I reinstalled Windows 10. And for a while, life was good and no blue screens occurred. The computer was installed using the UEFI Bios option this time around. This time I decided to install Windows with the Motherboard not overclocked in any capacity - it was set on default as if it came straight out of the box.

And just two hours ago, a BSOD occurred: Driver_Corrupted_EXPOOL. I reinstalled Windows, and it worked well for a while but it came back. The last thing installed was over two days ago (Logitech Keyboard). I need help. At this point, what do you think is the problem? Should I RMA the Motherboard? Should I RMA the CPU too? PSU? What could it possibly be?

The computer had the BSOD while no one was actively using it - it just occurred while idle. BlueScreenViewer says it was caused by ntoskrnl.exe.