Solved Random reboots, error reporting calls it BSOD, next gen hardware

[BotenRedWolf]

I'm going to post a few symptoms in here, as I'm not sure if they're related. I'll also post specs in the thread so you don't have to go dig them up in the dump files.

Machine:

Corsair CX-750
Asus Z97-A MB
Intel Core i5-4590
2x 8GB Kingston HyperX Fury DDR3-1600
2x GTX 760 (SLI)
2x Samsung EVO 250GB SSD (RAID0)
1x Seagate 2TB
1x LG BD-ROM/DVD-RW
3x AOC 21.5" IPS/LED

This is my first build in about 10 years, and I very well could have some sort of parts incompatibility that I'm not aware of.

Anyway, symptoms:

The first noticeable symptom is that in some newer games, specifically Watch Dogs, I can play for an indeterminate amount of time at incredible speeds, and then out of nowhere, the game will start to lag. Literally lag, not a drop in FPS exclusively, but ALL actions become slow. Sometimes it will recover after a minute or so, sometimes not until after a reboot.

The other one, and slightly more obnoxious: seemingly random reboots on desktop. This typically happens if the CPU is under heavy load. I used to do SETI@HOME work, don't anymore because of instability. I was attempting to encode some video earlier, and it also crashed.

If you need any more information, let me know what you need and I'll do what I can to get it to you.

I'm not even sure if the attached files will be relevant or not, but I hope so.

 

[essenbe]

I haven't had time to look at all of the files, but you have 2 dump files, both blaming the Graphics driver.

BugCheck 116, {fffffa8018cf6390, fffff8800fdd6d1c, ffffffffc000009a, 4}
*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
Probably caused by : [COLOR=red]nvlddmkm[/COLOR].sys ( nvlddmkm+98bd1c )
Followup: MachineOwner
---------

If you are overclocking the CPU or GPU, please set everything back to default values. Also, please go into BIOS and give me the values of the +12V, +5V and +3.3V readings.

Have you been monitoring the temps of your CPU and GPU especially during gaming or heavy work loads? if not olease start doing so. I believe Asus has a GPU utility that will do that as well as enable a custom fan profile to keep the temps lower. If you don't have a CPU temp program, I use Core Temp to monitor CPU temps.

I may suggest you reinstall the drivers by going to Nvidia.com and using the latest WHQL version. I believe there is a newer one than what you have. However, I don't believe the version is causing your crashing problems. Try using this tutorial to reinstall the drivers. http://www.sevenforums.com/tutorials/327414-nvidia-drivers-avoid-problems.html If you are using 3 monitors you may use the 3D vision drivers. The point is, if you don't use it, don't install it. But the procedure will ensure you of the cleanest install you can get.

Let us know how that works out for you. As far as your hardware I don't notice any incompatibilities at a quick glance. And, I'm sure you know, you are not the first person to have problems with watch Dogs.

If you would, please let us know the BIOS values and try the newer drivers, using the linked method and let us know if it makes any difference for you.

 

[BotenRedWolf]

Since I've got a triple display deal going on, I have the ability to leave monitoring programs running during play and watch them. I'll check the voltages and make another post. I'd check with AIDA when under load and when unloaded (older supplies could vary, I don't know about new ones...) but it won't show 12V.

I have nothing overclocked, all BIOS settings are at defaults except for the SATA is configured for RAID0.

 

[essenbe]

The BIOS is probably the most important one and you will get more accurate results from bios. But, you can't get the bios results while playing, so we have to rely on software. Just be sure to have only one monitoring program at a time running. Sometimes when 2 programs are reading the same sensor, you will get inaccurate results.

 

[BotenRedWolf]

Well, I was attempting to make a new post with the information as well as what was going on while game was launched with AIDA... and it blacked out and rebooted, so here's another dump zip thing.

Also, my PSU rail voltages for 3.3, 5, and 12V respectively: 3.312V, 5.120V, and 12.192V

3.3V dropped to about 3.26-3.28 right before it rebooted on me, not sure if that has any consequence or not...

 

[essenbe]

That report said the exact same thing as the other 2 did. I don't know why your 3.3V dropped, but even with the drop, it was still within ATX standards. Try to update the driver going by that tutorial and see if that makes a difference. How old is this build, and did you have this problem from the start? is there any activity you can do that will make it BSOD right away? Let me know and I'll get back with you tomorrow, or someone else may jump in shortly.

 

[BotenRedWolf]

The only way I can think of forcing it to act strange is to 'hyper-load' it, maxing out the cards. Used to do it with BOINC and SETI@HOME set up to use all card's power. It doesn't act up on older or less demanding games, and incredibly rare that it will do it when I'm not around and the machine isn't loaded down. The only thing that's weird is that I was encoding a video earlier, and I believe that uses CPU only, and it rebooted then too.

The build is about 2 or so weeks old, as the Z97-A board is about a month old and the 4590 only came out a couple weeks ago, I was actually the first one in the city here (who bought from local stores) to get one. I've been having the issue since hour one, even when I had a GTX 6xx instead. The hardware is all so damn new, could this be attributed to the age of the hardware?

 

[Golden]

In this case, perhaps we can isolate things better by running Driver Verifier.

Please do the following:

Run Driver Verifier for 24 hours or the occurrence of the next crash, whichever is earlier.
http://www.sevenforums.com/tutorials/101379-driver-verifier-enable-disable.html

Driver Verifier will cause your computer to run very sluggishly - this is normal. What it is trying to do is force your system to BSOD and isolate the offending driver/s. When it does, reboot, disable driver verifier, reboot as normal and upload the new dmp file/s here.

I recommend creating a system restore point before turning on driver verifier:
http://www.sevenforums.com/tutorials/697-system-restore-point-create.html

If your system fails to boot to desktop once driver verifier is enabled, turn it off by booting into Safe Mode:
http://www.sevenforums.com/tutorials/69585-safe-mode.html

 

[essenbe]

If you mean the new technology, I think not. Usually new technology has some issues until they have had time to work out the problems with bios updates and driver issues, but it is rarely to this extent. I know I went through it with my system, but it was usually bios settings didn't set to what you set them to and a lot of software hadn't caught up to the newer chipsets and things such as that. But, it usually does not involve BSODs. Your problem thus far seems to be graphics related. All of your dump files point to the graphics driver, which can also be the card itself too. Have you received any messages that the driver has crashed and recovered? The card is not that new and the drivers have had plenty of time to mature. I will check your board QVL list and see if you need a bios update for your CPU. The encoding is usually pretty stressful on the CPU, but depending on exactly what you were doing, could put some of the work on the GPU too. Have you run all of the Windows updates? if not, check again and see if there are more updates for your rig. Also, when you first installed the motherboard drivers, did you do it from the driver DVD or did you download them from Asus web site? The web site will have more up to date drivers than what was on the disk.

I have never run SLI, but in my experience with Graphics problems, it comes down to the Drivers, PSU, the board or the card itself. It is possible one of your cards are bad. We'll try some other things before we start pulling cards though.

 

[essenbe]

I checked the QVL list and a bios update is not required for your CPU. If you would, please download CPUz, if you don't already have it, and post screenshots of the CPU, mainboard, memory and SPD tabs. The SPD tab is listed by dim slots and can be selected which slot from a drop down list in the upper left portion of the window. I only need the dim slots that are populated with ram.

I'll post more tomorrow when you have tried the things I and Golden have suggested.

 

[BotenRedWolf]

I receive "driver has failed and recovered" messages on numerous bases. I will screenshot those things from CPUz in a little bit, since it also came with my board's software. I installed drivers off of the ASUS website, since I didn't have an optical drive in my build at the time it went together.

 

[BotenRedWolf]

Attached is a screenshot and a link to a full CPU-Z report in HTML form.
http://wavemadison.net/beastreport.html

I performed the nVidia driver special installation thing, so I'll run golden's driver crash test.

 

[BotenRedWolf]

I know this is damn near post farming, but I ran the driver verifier and it didn't even boot, straight to BSOD and instant reboot... twice... had to safe mode to disable verifier and now it works as it did before. So I'm uploading another diag report for y'all to check out.

Also, another thing popped into my head and I don't remember if I mentioned it before, but even when I had my old single GTX 670 or whatever it was, it crashed and acted funny like this too.

 

[essenbe]

Was the 670 in this machine or another machine? If it was another machine, what hardware from that machine was carried over to this machine? Also, Please go to Nvidia's web site to download the drivers. You will get more current drivers there, just install the latest WHQL drivers.

There were no new dump files in what you uploaded, the same ones as before.

 

[BotenRedWolf]

670 was in this machine, used it for like a week, and realized it wasn't going to keep up with me so I updated... hard...

I also did a fresh driver download when I followed the nVidia driver instructions

 

[essenbe]

Try something for me. Download Ungine Heaven 4.0 (Download Unigine 'Heaven' DX11 Benchmark 4.0 | techPowerUp) just watch out for the add ons they try to install with it. Set Heaven up like this, just select how you run the monitors in the multi monitors section, when you start it , at the top click Benchmark. But before you click run, do this http://www.sevenforums.com/tutorials/330579-graphics-card-problems-diagnose.html. It will create a log which will be quite long, so as soon as you set up the log, start Heaven and as soon as Heaven stops, uncheck the log to file box to stop it.

View attachment 320239

 

[BotenRedWolf]

Okay. I tried with Unigine (which I already had), as well as SETI@HOME (which almost always caused some shit), and Watch Dogs. Unigine I did with wall 3x1 and single display. None of them would do anything except stutter occasionally in watch dogs, which is understandable since the game is so damned new and probably not optimized quite right yet.

 

[essenbe]

I really couldn't see anything in the log files. I'll look again and see if I can see anything. How do they work in less demanding games than Watch Dogs?

 

[BotenRedWolf]

Absolutely phenomenal, as expected. I'm wondering if this was a glitchy driver thing, since now it seems to be working, for the most part, as it should be...

 

[essenbe]

What did you do to fix it?