problems with 4th gen i7 4770 and 1080ti

[castortroysi]

So, I've been upgrading my system and building a few systems for others for a long time. But I have to reach out over my new mystery problem. The system is not the latest greatest I know, but its up there.

OS: Windows 10

Motherboard: Asus Z97 Deluxe
https://www.newegg.com/Product/Product.aspx?Item=N82E16813132127

CPU: i7 4770k
https://www.newegg.com/Product/Product.aspx?Item=N82E16819116901

Cooler: Corsair H100

RAM: Corsair Vengeance Pro 16GB (2 x 8GB) 2400mhz
https://www.newegg.com/Product/Product.aspx?Item=N82E16820233585

GPUs(1 at a time):
(stable)EVGA GTX 780 SC
https://www.newegg.com/Product/Product.aspx?Item=N82E16814130918

(unstable)EVGA GTX 1080ti FTW3
https://www.newegg.com/Product/Product.aspx?Item=N82E16814487338

Zotac GTX 1080ti Founders Edition
https://www.newegg.com/Product/Product.aspx?Item=N82E16814500415

PSU: Corsair HX1050
https://www.newegg.com/Product/Product.aspx?Item=N82E16817139034

SSD: Cosair XTi 1tb

Case: Corsair 600T

Monitor: 60hz 8ms 1080p

The issue is hard restarting during gaming. It all started when I picked up a founders edition 1080ti to replace my evga 780. Swap in the new card and immediately start playing my favorite titles that tend to use a lot of the system resources. Once the system warms up for a good 15min or so suddenly it decides it needs to restart on its own. To better describe the symptoms: it seems to be a heat soaking issue due to it needing to really warm up and require the newest highest resource demanding games to cause the hard restart to occur. But, sometimes it will play those same games for hours, days even without an issue(ambient temps could be part of this?). Other times, it will be a cold night with cold ambient temps and the crashing/restarting occurs. The pc case remains open at all times and I've even tried putting a box fan in front of it to no avail. Yet logging temps at time of crash shows no high temps.

I've definitely gone through the motions here with memtest, reinstalling the OS, underclocking the memory to 1600, updating mobo bios, updated the 1080ti bios, underclocking the GPU, setting nvidia control panel to prefer maximum performance, setting the windows power profile to use 99% cpu power, logged temps during crash(<60c for cpu, <70c for gpu), tried different power rails on the PSU, switched back to the 780 to find again my perfect stability, swapped the founders edition 1080ti for a ftw3 1080ti, yet the problem persists.

Before you jump to point to the PSU, the specs are showing a 250w on the 780 and 20-30w LESS on the new 1080ti's. And the system is stable when I install the 780 back in.

So, really I need to locate the cause of my problem here. I don't have money to throw at it or many spare parts. I do have a second system here if I really need something else to test with but I am looking for the smart ways to find this problem. I do have a multimeter if needed.

There are only so many things that are allowed to trigger a hard restart. Sensors throughout the boards.... How do I identify which one is causing this forced hard reset?

Is it an issue with the CPU and RAM compatibility? Even though the RAM is 2400mhz the mobo is 1600mhz or "2400mhz(OC)" and the CPU is not overclocked still at 3.5... although I have tried 3.9 with the exact same crashing/restarting results.

Could it be something else on the motherboard causing such a behavior?

PLEASE HELP this is such a drag on daily operations and general happiness. Let me know if I can answer any more questions or do any specific tests so we can solve this.

Did I miss something when reviewing the logs?
here is HWinfo diag: http://pasted.co/7a893561

here is log example of when it crashed at 54c(GPU) and 37c(CPU) when going back to the menu of a game and earlier in that game reached the targeted 83c(GPU) and the CPU was at 43c: http://pasted.co/a6bcff7e

***edit - Don't run 2400 ram in a mobo meant for 1600 or you may burn out your CPU

 

[castortroysi]

Today, I was able to play PUBG most of the day no problem, but eventually had a crash. Also, was able to run Superposition benchmark "game" mode with wireframes on and moved all of the objects to float around in 8k resolution dropping the FPS to around 22 for a couple of hours no problem. Although, yesterday it crashed when I was in the "game" mode and first started making objects float around. Any suggestions welcomed here. I'm willing to try new things or do things again to diagnose or test the issue.

 

[mdd1963]

Once could end up with mysterious sets of symptoms if a mainboard is unable to supply sufficient power thru the PCI-e slot, where one GPU works, another does not, etc...

However, the PSU is still a suspect until eliminated. You'd also need the 1080Ti GPU tested in another rig (equipped with a sufficient PSU) just to rule it out...

 

[castortroysi]

Hmm... any way to detect said mainboard PCI-E power issue?

Tried an entirely new 1080ti(from founders edition to ftw3) with same problem only slightly lower temps and less crashing. Also, tested them in another system with no stability issues.

Symptoms really seem like heat soaking. I just don't see how there could be a hard switch to shut a system down but no sort of log or notification of which switch was tripped. Does anyone know how to find this info? As you can see from the original post, it is not showing in the HWinfo log.

 

[ameyer75]

Hmm... any way to detect said mainboard PCI-E power issue?

Tried an entirely new 1080ti(from founders edition to ftw3) with same problem only slightly lower temps and less crashing. Also, tested them in another system with no stability issues.

Symptoms really seem like heat soaking. I just don't see how there could be a hard switch to shut a system down but no sort of log or notification of which switch was tripped. Does anyone know how to find this info? As you can see from the original post, it is not showing in the HWinfo log.

I agree with mdd1963. If you haven't ruled out your PSU, you should do that first, especially if yours is dying. It could discharge and take the rest of your computer with it. If you swap out the power supply and it makes no difference, you can return the new power supply and then the motherboard is next. Just swap the PSU and try it out for a few days, then let us know what happens.

 

[castortroysi]

I don't want to buy a PSU to test with. I have a multimeter and the sense to use it... Still need to find where to check and what readings I'm looking for but how does it make sense to have a stable system with a video card that uses MORE power? Furthermore, this does not address the symptom change of less crashing with a 1080ti that has better cooling(and perhaps more power usage with the RGB LEDs and the higher clock speeds)... Heat seems to be the issue here and its not effecting the PSU as it is a 1kw for a system only need half of that, with an open case and plenty of breathing room for the PSU being on a hard surface, not carpet.

780 = stable system
1080ti founders edition = lots of crashing and temps for GPU reading in the 80c, with some crashes occuring with a 50c reading
1080ti FTW3 = occasional crashing, once or twice a day after long sessions and temps never reaching above 65c

 

[mdd1963]

There is no real adequate tester for PSUs under a heavy load other than a heavy load as you are subjecting it to....

If you get crashes with 2 different 1080's, then your PSU is VERY a possible suspect, and one most would prefer to change before blaming the mainboard....

If it is heat soaking issue, then it hypothetically will not appear when an accessory fan is blown into the case with the case opened up....

 

[Karadjgne]

I'd start with updating all of the mobo drivers, not just the bios. Lan, USB controller, audio, Sata controllers etc. I'm not so sure it's a power issue as much as driver conflict, windows 10 updates have issues with resetting defaults, and out-of-date drivers, replacing them with newer versions of windows generics.

 

[castortroysi]

I revisited the PSU, the fan was down-facing with a filter that was NOT clogged with dust and a hard surface with room to breath. I turned it over with the fan opening facing upwards. My theory here is that the new GPU opens the previous bottleneck and allowed the CPU to use more power generating more heat at the PSU. It was indeed hotter to the touch than ever before. Still crashed! At time of crash, no excessive heat blowing out of the PSU or to the touch this time. I'm getting a loaner PSU from a friend and I will update tomorrow. Meanwhile, I will try examining the drivers again, although, I did reinstall windows for this issue and at the time visited the ASUS site to obtain the up-to-date drivers. Stay tuned!

 

[castortroysi]

I revisited the ASUS site for drivers and they all downloaded with (1) at the end... the same after extracting them, and when running the setups - prompted me to repair or uninstall each. Reviewed them in device manager and all looked clear. There are a few extra microsoft drivers that I don't fully understand but I wouldn't know just by looking at them if they are supposed to be microsoft or not. The only one that stood out was the drop down for "processors" that had each of the 8 cores(4 of those virtual obviously) and they were all microsoft drivers. Intel chipset driver was installed so I dunno what that's all about or if it has anything to do with my issue. Test PSU going in tomorrow.

 

[castortroysi]

New PSU results in the exact same symptoms.

 

[Karadjgne]

Have you browsed windows event viewer? Critical errors causing restarts/bluescreens are usually listed under the admin profile.

 

[castortroysi]

The event viewer is where I started. I do have some stuff in there but nothing related to this issue. Revisiting the log now, some days that the pc will crash 1-4 times and there are no events what-so-ever. I'm trying the RAM from the donor computer now to help with our process of elimination. Stay tuned!

 

[castortroysi]

Ok, that crashed, too. Not the RAM. To clarify, there is no bluescreen at all. Only hard restarting itself when stressed to the max using the most resource intensive games and benchmarking software. I can run some games that are less intensive all day without crashing.

 

[castortroysi]

Something else worth mentioning is that my system has always been protected by a very nice power conditioner that has a $5 million connected-equipment warranty. https://www.crutchfield.com/S-F3bD883wWtK/p_299M5300EX/Panamax-MAX-5300-EX.html

But not a battery backup.

 

[castortroysi]

Another thing to note is that the fan speed profile on the GPU is a rather steep curve reaching 90% fan speed by 70c and 100% by 74c. With the current setup with the FTW3, usually the temps don't go over 65c. Strangely, it will take 15-20min with this curve before it will crash during gaming.... yet, if I set the fans to 100% before starting the game and leave them there, it can last an hour or two before crashing.

 

[mdd1963]

"Something else worth mentioning is that my system has always been protected by a very nice power conditioner"

Skip the UPS during testing, since your system is unstable....; it'd be a shame to not have ruled out that simple input source MUCH sooner....

 

[mdd1963]

Test for stability with your 1080Ti clocks at more 'reference' levels of gpu /mem speeds....

 

[castortroysi]

Um, is English your first language? I'm having trouble understanding what you mean by that. The point I was trying to make is that the system has clean power and always has, even though the power in the structure has had brown- and black-outs. Basically it has had a really nice power surge protector, but has lost power during usage at times since there is no battery backup.

The power cleaner is not an issue... Its over engineered to the point the manufacturer is willing to warranty $5m in your electronics that you plug into it, and at the time of my PC's crashing no other device plugged in experiences any sort of outtage. Furthermore, the PC has been plugged into alternate outlets on the fancy power surge protector.

 

[castortroysi]

The first 1080ti was a reference model and was tested with an underclock as well. Even alternate versions of the firmware were flashed for testing, all with the same crashing.

 

[castortroysi]

Wow, I've reverted back to my 780 for stability. 1080ti paperweight lol

 

[mdd1963]

U

The power cleaner is not an issue... Its over engineered to the point the manufacturer is willing to warranty $5m in your electronics that you plug into it, and at the time of my PC's crashing no other device plugged in experiences any sort of outtage. Furthermore, the PC has been plugged into alternate outlets on the fancy power surge protector.

"Power Cleaner" ....Are you running off a fancy passive surge suppressor, or an active battery back-up/UPS? If an UPS, do you think they keep their pristine '750 watt' ratings, or whatever, forever? A warranty for "X millions of dollars of equipment' is not the same thing as supplying effective power. Your system is suffering crashes, and If you do not wish to try alternate simple non-UPS types of 115VAC sources to at least eliminate simple input power as a potential cause.... well....replace everything else expensive 3 times before ruling out the simple, i guess....

 

[castortroysi]

I don't know how else to explain that there is NO BATTERY BACKUP than "there is no battery backup," "Basically it has had a really nice power surge protector," and posting a link to the full description with detail regarding the power conditioner that shows, it does not have a battery.

 

[mdd1963]

It's possible your motherboard is insufficient to accommodating the 1080's power demand, for whatever reason...

(Since you refuse to even *consider* skipping your fancy $15-$30 dollar "Power Cleaner", as though trying a different outlet on the same filtering device has somehow eliminated it from consideration, .....I suggest a new motherboard...; if that fails, perhaps then you'll consider that a 'power conditioner' might possess filtering circuitry not *quite* allowing the rapid current increases required by a PSU supplying a GPU ramping up from 40 watts to 200 extra watts in power nearly instantaneously...

Since skipping the "power cleaner' would take a full 1-2 minutes of effort, by all means...proceed with another mainboard or GPU swap.)

 

[castortroysi]

Will you please read the whole thread before replying? at this point you're just filling it up with junk making it more difficult for people to finish reading so that we can eventually get to a solution.

Several post ago I wrote: "Furthermore, the PC has been plugged into alternate outlets on the fancy power surge protector."

This power conditioner(retail is $500) is running a lot of equipment, including a second gaming PC and I assure you, it is of no concern. It was brought up to prove the equipment has had clean power its entire life and to admit there is no battery back up, resulting in a few sudden crashes from power outages. Maybe this does or doesn't have an affect, but the expert electrical engineers will know and I wanted to include these details in explaining the entire system and those factors that may influence this problem I'm having.