©ASUSTEK COMPUTER INC. ALL RIGHTS RESERVED.

enyceedanny · ‎10-06-2016

I've recently upgraded to the 6850K and Rampage V Edition 10. After much testing, I found at 44x multiplier at 1.35v runs everything stable (Prime95 26.6 and latest, AIDA64), but during Realbench runs Luxmark would sporadically crash, with an event viewer indicating that the nvidia driver crashed then recovered.

After a lot of searching around, I found that certain monitoring software can cause it to crash. Although it was generally only with the OSD part on. But even with no monitoring software running, it would still crash frequently.
I've even put my CPU back to stock and downclocked my DIMMS to 2133mhz with loose timings only to have the display driver crash/recover during realbench. This didn't only happen on Realbench, it would happen during LINX as well (but not P95).

At this point I thought maybe it was my video card, but it passes any and all stability testing I threw at it.

For the hell of it, I decided to up the voltage of my DIMMS to 1.36 (from 1.35 stock), and what do you know, it doesn't crash anymore. I've been running additional tests throughout, but so far it hasn't crashed yet. Even back at my overclock settings with tight timings.

When looking at the Qualified Vendor List for DDR4, my set isn't listed there, so I'm guessing the BIOS isn't really optimized for them. I have the Corsair Dominator Platinum 2800 (4x16gb) modules. I'm guessing that for higher density DIMMS, a slight notch in vdimm is needed for full stability.

Although YMMV, if you're having such issues when everything else seems to be stable, try it out. :rolleyes:

Oh, and one way to replicate the crashing easily, is to have RB running stress test @ max memory, and have a youtube video running fullscreen. Then keep getting in and out of fullscreen mode. It will start to lag up the whole system for a bit, then the display driver will crash. It may take a few minutes, but I've been able to replicate it like this every time.

Qwinn · ‎10-07-2016

Were you running in SLI? Realbench is known to crash with SLI enabled during the Luxmark portion (and also, in my experience, while gathering system information). IMHO the fact that Realbench yields "false positive" crashes when using a very common configuration like SLI devalues Realbench a great deal for me as a proper stability test.

enyceedanny · ‎10-07-2016

Qwinn wrote:
Were you running in SLI? Realbench is known to crash with SLI enabled during the Luxmark portion (and also, in my experience, while gathering system information). IMHO the fact that Realbench yields "false positive" crashes when using a very common configuration like SLI devalues Realbench a great deal for me as a proper stability test.

Nope, single GPU.

Arne_Saknussemm · ‎10-07-2016

Hmmm...errrmm...hang on. I run SLI and have no problem running RealBench and never have...on any version or any Nvidia driver...so that seems too general a statement to me...

I reproduced an issue with OSD running from Precision X on a previous version...since I don't use OSD a lot I don't have it on much and haven't tested latest version...

This is the only "false positive" I have ever had from RealBench...and even then I'm not sure I could categorically use that description. Crashes in RealBench are crashes and come from instability in 99.999% of cases. As OP kind of attests...

I have had problems with Luxmark not starting sometimes but clean install of Nvidia drivers fixed that...

ARNE’S GUIDE TO PC BUILDING HAPPINESS

ARNE'S GUIDE TO OVERCLOCKING HAPPINESS

Qwinn · ‎10-07-2016

Here's a thread where several people get black screens running the benchmark in SLI that go away when SLI is disabled:

http://rog.asus.com/forum/showthread.php?74446-realbench-crash-when-the-stress-test-is-stopped/page2

That's the second page of the thread. Note near the end that I did extensive testing with running on all optimized defaults, no overclocking of any kind, with SLI enabled and it would still crash.

Arne_Saknussemm · ‎10-07-2016

I would just assume SLI is adding more "stress" to system and RealBench is finding weakness at that stress level. That would honestly be my first take. As I say I have no problem running RB with SLI and SLI OCd and everything OC'd....up to a certain point of course.

Of course there may be some weird problem/conflict with RB in some cases that you guys have found but that would be my last thought TBH. I'd look at OC or mem stability first...

ARNE’S GUIDE TO PC BUILDING HAPPINESS

ARNE'S GUIDE TO OVERCLOCKING HAPPINESS

cynique · ‎12-08-2016

enyceedanny wrote:
I've recently upgraded to the 6850K and Rampage V Edition 10. After much testing, I found at 44x multiplier at 1.35v runs everything stable (Prime95 26.6 and latest, AIDA64), but during Realbench runs Luxmark would sporadically crash, with an event viewer indicating that the nvidia driver crashed then recovered.

After a lot of searching around, I found that certain monitoring software can cause it to crash. Although it was generally only with the OSD part on. But even with no monitoring software running, it would still crash frequently.
I've even put my CPU back to stock and downclocked my DIMMS to 2133mhz with loose timings only to have the display driver crash/recover during realbench. This didn't only happen on Realbench, it would happen during LINX as well (but not P95).

At this point I thought maybe it was my video card, but it passes any and all stability testing I threw at it.

For the hell of it, I decided to up the voltage of my DIMMS to 1.36 (from 1.35 stock), and what do you know, it doesn't crash anymore. I've been running additional tests throughout, but so far it hasn't crashed yet. Even back at my overclock settings with tight timings.

When looking at the Qualified Vendor List for DDR4, my set isn't listed there, so I'm guessing the BIOS isn't really optimized for them. I have the Corsair Dominator Platinum 2800 (4x16gb) modules. I'm guessing that for higher density DIMMS, a slight notch in vdimm is needed for full stability.

Although YMMV, if you're having such issues when everything else seems to be stable, try it out. :rolleyes:

Oh, and one way to replicate the crashing easily, is to have RB running stress test @ max memory, and have a youtube video running fullscreen. Then keep getting in and out of ful lscreen mode. It will start to lag up the whole system for a bit, then the display driver will crash. It may take a few minutes, but I've been able to replicate it like this every time.

I have a similar problem my graphics card is overclocked but i also tried at stock clock speeds but realbench either crashes as soon as i run it or after 5-10 minutes lexmark crashes.Gpu seems stabil overclock i tested it with unigine and firestrike.I think gskill trident z 16 gb rams(2x8) may be the problem but xmp on and off no luck.Cpu(6700k) not overclocked right now but doesnt make a difference if i oc it.Still RB crashes

enyceedanny · ‎12-08-2016

cynique wrote:
I have a similar problem my graphics card is overclocked but i also tried at stock clock speeds but realbench either crashes as soon as i run it or after 5-10 minutes lexmark crashes.Gpu seems stabil overclock i tested it with unigine and firestrike.I think gskill trident z 16 gb rams(2x8) may be the problem but xmp on and off no luck.Cpu(6700k) not overclocked right now but doesnt make a difference if i oc it.Still RB crashes

Although the bump in vdimm did help initially, the issue came back. In the end, a clean install of Windows 10 solved this.

Avenger411 · ‎12-09-2016

Hey,

Having an OSD also triggers this kind of behaviour. Just so you know, arne is right.

Max

Cpu : Intel 5930K@4.25ghz@1.2v / Cache @4.25ghz@1.20v
Cpu Cooler : Corsair H100i
Case : Corsair 780T
Memory : G.Skills 32GB DDR4-3200mhz CAS 15-15-15-35-1T@1.370v
Motherboard : Asus Rampage V Extreme (BIOS 3504)
M2 : Samsung 950 Pro NVME 512gb (Gaming)
M2 : SSD1 : OCZ RD400A 128gb (windows)
SSD1 : Crucial MX100 512gb (data)
Gfx : EVGA Titan X w/ 980 Hybrid Cooling AiO Liquid Cooler
PSU : Antec HCP-1000W
Monitor : Asus RoG Swift

xPepi · ‎01-21-2018

I also had luxmark.exe crash during stress testing and I did as someone suggested by closing all gpu related applications...afterburner, rivatuner, oc guru etc...closed it all, disabled screen saver and screen shut off and voila...no crash, ran stress test for an hour.

Potential Fix. Realbench Luxmark has crashed + display driver crash/recovered.