Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] nVIDIA GeForce GTX TITAN X Owners Club
New Posts  All Forums:Forum Nav:

[Official] nVIDIA GeForce GTX TITAN X Owners Club - Page 1770

post #17691 of 18250
I'm willing to be this is related to temps. Leave Open a monitoring software of your choice and keep it running in the background. I had something similar to what you are explaining happen to me previously. Even though my card is on water is seems right at the point the card hit a certain temp is would crash when loaded hard again. This could very wel be what you are explaining and why it appears so random. Try increasing the pump rpm a bit and or fans and see if that eliminates it. Even reducing the OC may not help as these cards can be very picky with temps from my experience.


Always destroying exergy
post #17692 of 18250
Quote:
Originally Posted by superkyle1721 View Post

I'm willing to be this is related to temps. Leave Open a monitoring software of your choice and keep it running in the background. I had something similar to what you are explaining happen to me previously. Even though my card is on water is seems right at the point the card hit a certain temp is would crash when loaded hard again. This could very wel be what you are explaining and why it appears so random. Try increasing the pump rpm a bit and or fans and see if that eliminates it. Even reducing the OC may not help as these cards can be very picky with temps from my experience.


Always destroying exergy


I always have HWinfo running

max GPU temp is 51-60c max depending on ambient

what temps was yours crashing at?
Code55
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-6700k 4.8 Asus Impact VIII / EK M8i Monoblock SeaHawk EK X 1080 Ti G.Skill Ripjaw V 2 x 16GB 3733CL15 1T 
Hard DriveHard DriveCoolingMonitor
INTEL 750 400GB U.2 SanDisk Extreme Pro 960GB EK D5 PWM Acer X34A / XB321HK 
KeyboardPowerCaseAudio
Logitech G810 EVGA 750 G2 In-Win D-Frame Mini Meridian Explorer 
  hide details  
Reply
Code55
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-6700k 4.8 Asus Impact VIII / EK M8i Monoblock SeaHawk EK X 1080 Ti G.Skill Ripjaw V 2 x 16GB 3733CL15 1T 
Hard DriveHard DriveCoolingMonitor
INTEL 750 400GB U.2 SanDisk Extreme Pro 960GB EK D5 PWM Acer X34A / XB321HK 
KeyboardPowerCaseAudio
Logitech G810 EVGA 750 G2 In-Win D-Frame Mini Meridian Explorer 
  hide details  
Reply
post #17693 of 18250
Quote:
Originally Posted by Asus11 View Post

thanks for the input, just took the CPU overclock off but didn't help frown.gif

Take this for whatever it's worth, but I had lots of display driver crashing when Game DVR was enabled in Windows 10. I turned it off and the problem went away entirely. I too thought it was a bad overclock, then worried my GPU was acting up after custom BIOS / additional voltage but..nope. All software.

Quote:
Originally Posted by Asus11 View Post

my titan x just keeps crashing when overclocked

could be stable 3 hours

then randomly will driver crash wth!

no idea how to actually find a stable overclock .. considering its so random and unpredictable

I have changed my opinion on it based on real world results with both GPU Boost 2.0 and boost disabled with custom VBIOS / max input voltage at the GPU core.

Real world, not a lot of discernable difference. Benchmarks? Definitely an improvement there. I don't play benchmarks, though (I use them to compare to a baseline, 99% of the time, which is really where I find them most useful for those who aren't into competitive overclocking.)

So what I'm doing now is this: I have an EK Waterblock on mine, which eliminates the thermal throttling issue completely.

I have the EVGA Hydro Copper BIOS with the 425 Watt power limit. See the first post in this forum.

I set the power limit to 120%.

I leave the Nvidia control panel option for power at "Adaptive" and my Windows power profile is also at Balanced. I've run long enough with them set at "Maximum Performance" and "High Performance" to see that the tradeoff is good enough *when I'm not running a benchmark.* -I live and work from home in Texas, and around 4-5pm I generally have to evacuate my western facing office due to the temperature increase from the setting sun. I need any reduction in room temperature that I can get.

Other than that, I don't touch the GPU or memory clocks. I tested 8000 MHz memory, and it didn't crash, but again, I can't really tell much of a difference between standard and manually overclocked when playing real games at 1440p. Since I'm not developing CUDA software or using the GPU as a compute engine, I'm not really losing anything here in probably 95-99% of the cases (the games I play.)

Yes there can be 5-10fps difference depending on settings, comparing stock to max OC, but I have a GSync monitor, and something like The Division runs great at 1440p / Ultra on a TITAN X. I'm usually between 72 and 100 fps on that title with close to max settings.

What about results? Well, with the way I have described my config, GPU Boost will go up to 1342 MHz and overall power draw is around 340 Watts (on average.)

Compared to the ultimate BIOS with maximum allowable voltage it's only a 104 MHz difference in clock speed.

Considering Maxwell has a very steep rate of diminishing returns (and Pascal seems to have this even more so at the new process node) I'm comfortable with that!
Edited by mypickaxe - 6/6/16 at 11:08pm
post #17694 of 18250
Quote:
Originally Posted by Asus11 View Post

Quote:
Originally Posted by superkyle1721 View Post

I'm willing to be this is related to temps. Leave Open a monitoring software of your choice and keep it running in the background. I had something similar to what you are explaining happen to me previously. Even though my card is on water is seems right at the point the card hit a certain temp is would crash when loaded hard again. This could very wel be what you are explaining and why it appears so random. Try increasing the pump rpm a bit and or fans and see if that eliminates it. Even reducing the OC may not help as these cards can be very picky with temps from my experience.


Always destroying exergy


I always have HWinfo running

max GPU temp is 51-60c max depending on ambient

what temps was yours crashing at?
Hey sorry for the late reply I was away. It's odd honestly I would get through firestrike and any other benchmark no prob. Some games were even stable also but playing a game such a the division in 4K would be fine for a good while. I would look at menus reducing clocks and when the load was applied again it would crash. Same on Witcher. Would only happen when load was reduced and reapplied and temps were above 50. I set a more aggressive fan profile on the rad and hasn't crashed since. For your case however I would suggest trying it. If it doesn't work then also try using DDU and doing a clean install of various drivers starting with the newest and slowly working down revisions. All drivers are not created equally. You should find one your card likes better than others and stick with that one for a while until you feel like testing the stability of some of the newer drivers.


Always destroying exergy
post #17695 of 18250
[/quote]
Quote:
Originally Posted by Asus11 View Post

my titan x just keeps crashing when overclocked

could be stable 3 hours

then randomly will driver crash wth!

no idea how to actually find a stable overclock .. considering its so random and unpredictable

How many applications / games is this in ? if its one that you are focussing on i would really probably ignore it for now and explore other applications. I was playing TW3 in 3D vision at clocks from 1600 mhz to Default in SLI and it would just randomly crash in TX SLI , sometimes 3 hrs hours in , sometimes 15 mins in. I never did solve it , but after 115 hrs of game time i've had enough of TW3 for the moment lol.

Lately i have not had the chilled rig up and running, no room at the new place for it so its been an ambient liquid cooled rig with TX SLI at 1525/8100 , 39-44c max temps - its all incase so that's about the best temps i can get at present. Played 22hrs of Tombraider so far and quite regularly 98-99% usage a lot of the time in 4k and its solid. Warhammer single card same OC is fine , 16 hrs in. Doom same OC 8 hrs in , all good. I think it would likely crash if i went a bit further at these temps but i'm getting 60 fps in 4k (most of the time) so no need for higher clocks at present.

So far one game seems problematic and that's AOTS in dx12 multi adapter. I need to just seem if the game in general is unstable in explicit adapter mod or if its the overclock. I need to force myself to play AOTS a bit more , but waiting on DX12 warhammer and vulken Doom ..

I'd keep a note pad on it, game/application - clocks - temps and how long you play for etc. You might notice a trend. If not you will just have it recorded to look back on , can be good if you doubt your memory etc.

One thing i have done , which i think is a good addition if you run a back plate is to add a fan to the back plate , this likely shaves a few C off your ram temps and PCB in general.
post #17696 of 18250
Quote:
Originally Posted by KedarWolf View Post

I use them on water but I don't have an IR gun to measure the back on the card with the block on. I just noticed when I installed each one the EK has more thermal pads on the VRMs and memory modules etc. The EVGA has more limited coverage.

But under water with a highly overclocked CPU and a Titan X with a 1.281v maxed out voltage bios I get under 75C CPU and under 55C video card.

Check out the Predator 360 with the Quick Disconnect (QDC) GPU blocks.

http://www.overclock.net/t/1566468/ek-predator-club-discussion-thread

https://www.ekwb.com/news/ek-pre-filled-and-qdc-enabled-gpu-water-blocks-are-finally-here/


Easy way to go water on CPU AND one video card. Just install the AIO, put on the EK backplate and GPU block, clip the video card water block on to the 360, problem solved!

It's what I use. biggrin.gif
Nice i seen these did not realize you could hook a GPU up to it..
I think i will wait for the Next Titan then upgrade my whole system to these EK Coolers so easy to install..
no messy custom water Loop

My Hybrid and H100 will have to do untill next year dont want to spend any more on my T-X with water block
Edited by romanlegion13th - 6/7/16 at 12:26pm
post #17697 of 18250
Quote:
Originally Posted by Asus11 View Post

my titan x just keeps crashing when overclocked

could be stable 3 hours

then randomly will driver crash wth!

no idea how to actually find a stable overclock .. considering its so random and unpredictable

That's just what overclocking has been the last decade or so? smile.gif

I found Witcher 3 and Fallout 4 crashing sometimes but not even every night. Just an unstable overclock. I am running my TXs at 1455 MHz at 1.274 V currently. VRAM not overclocked. Very stable, not a single crash since I dialed in those clocks so I am getting close to final stability in games I play.

Also I would try without any 3rd party apps like Precision or Afterburner if you are using those. Some games crash only because they are in the background, Witcher 3 for example. Just mod the BIOS to your needs...thumb.gif
post #17698 of 18250
Quote:
Originally Posted by mypickaxe View Post

Take this for whatever it's worth, but I had lots of display driver crashing when Game DVR was enabled in Windows 10. I turned it off and the problem went away entirely. I too thought it was a bad overclock, then worried my GPU was acting up after custom BIOS / additional voltage but..nope. All software.
I have changed my opinion on it based on real world results with both GPU Boost 2.0 and boost disabled with custom VBIOS / max input voltage at the GPU core.

Real world, not a lot of discernable difference. Benchmarks? Definitely an improvement there. I don't play benchmarks, though (I use them to compare to a baseline, 99% of the time, which is really where I find them most useful for those who aren't into competitive overclocking.)

So what I'm doing now is this: I have an EK Waterblock on mine, which eliminates the thermal throttling issue completely.

I have the EVGA Hydro Copper BIOS with the 425 Watt power limit. See the first post in this forum.

I set the power limit to 120%.

I leave the Nvidia control panel option for power at "Adaptive" and my Windows power profile is also at Balanced. I've run long enough with them set at "Maximum Performance" and "High Performance" to see that the tradeoff is good enough *when I'm not running a benchmark.* -I live and work from home in Texas, and around 4-5pm I generally have to evacuate my western facing office due to the temperature increase from the setting sun. I need any reduction in room temperature that I can get.

Other than that, I don't touch the GPU or memory clocks. I tested 8000 MHz memory, and it didn't crash, but again, I can't really tell much of a difference between standard and manually overclocked when playing real games at 1440p. Since I'm not developing CUDA software or using the GPU as a compute engine, I'm not really losing anything here in probably 95-99% of the cases (the games I play.)

Yes there can be 5-10fps difference depending on settings, comparing stock to max OC, but I have a GSync monitor, and something like The Division runs great at 1440p / Ultra on a TITAN X. I'm usually between 72 and 100 fps on that title with close to max settings.

What about results? Well, with the way I have described my config, GPU Boost will go up to 1342 MHz and overall power draw is around 340 Watts (on average.)

Compared to the ultimate BIOS with maximum allowable voltage it's only a 104 MHz difference in clock speed.

Considering Maxwell has a very steep rate of diminishing returns (and Pascal seems to have this even more so at the new process node) I'm comfortable with that!

Quote:
Originally Posted by superkyle1721 View Post

Hey sorry for the late reply I was away. It's odd honestly I would get through firestrike and any other benchmark no prob. Some games were even stable also but playing a game such a the division in 4K would be fine for a good while. I would look at menus reducing clocks and when the load was applied again it would crash. Same on Witcher. Would only happen when load was reduced and reapplied and temps were above 50. I set a more aggressive fan profile on the rad and hasn't crashed since. For your case however I would suggest trying it. If it doesn't work then also try using DDU and doing a clean install of various drivers starting with the newest and slowly working down revisions. All drivers are not created equally. You should find one your card likes better than others and stick with that one for a while until you feel like testing the stability of some of the newer drivers.


Always destroying exergy

Quote:
Originally Posted by Orthello View Post

How many applications / games is this in ? if its one that you are focussing on i would really probably ignore it for now and explore other applications. I was playing TW3 in 3D vision at clocks from 1600 mhz to Default in SLI and it would just randomly crash in TX SLI , sometimes 3 hrs hours in , sometimes 15 mins in. I never did solve it , but after 115 hrs of game time i've had enough of TW3 for the moment lol.

Lately i have not had the chilled rig up and running, no room at the new place for it so its been an ambient liquid cooled rig with TX SLI at 1525/8100 , 39-44c max temps - its all incase so that's about the best temps i can get at present. Played 22hrs of Tombraider so far and quite regularly 98-99% usage a lot of the time in 4k and its solid. Warhammer single card same OC is fine , 16 hrs in. Doom same OC 8 hrs in , all good. I think it would likely crash if i went a bit further at these temps but i'm getting 60 fps in 4k (most of the time) so no need for higher clocks at present.

So far one game seems problematic and that's AOTS in dx12 multi adapter. I need to just seem if the game in general is unstable in explicit adapter mod or if its the overclock. I need to force myself to play AOTS a bit more , but waiting on DX12 warhammer and vulken Doom ..

I'd keep a note pad on it, game/application - clocks - temps and how long you play for etc. You might notice a trend. If not you will just have it recorded to look back on , can be good if you doubt your memory etc.

One thing i have done , which i think is a good addition if you run a back plate is to add a fan to the back plate , this likely shaves a few C off your ram temps and PCB in general.[/quote]


tested nearly all the bios on first page now

also already did the driver uninstaller and fresh driver install it didnt help

thinking what I noticed was with every single bios the card would scream as if it was on the end of valley credits

the best bios which it really likes is the 1256mv by sheyster it doesn't coil whine like a banshee and it keeps solid on it

also I do know what you mean previously the card would be stable for ages and ages then stuff like you said when you go from undemanding to demanding again it would crash like when I tab out for a sec then go back in etc

also tried the fixit not sure if it helps

anyway tested the card @ 1400/8000 was stable, now on 1450/8000 seems stable going to increase by 10 on core each day and see how far it will go

went into the bios and changed the PWM for pump/ fans slightly hoping for the best and put the cpu overclock back on

thanks everyone for the suggestions ill keep you updated with how I get on, hopefully good smile.gif



biggrin.gif
Code55
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-6700k 4.8 Asus Impact VIII / EK M8i Monoblock SeaHawk EK X 1080 Ti G.Skill Ripjaw V 2 x 16GB 3733CL15 1T 
Hard DriveHard DriveCoolingMonitor
INTEL 750 400GB U.2 SanDisk Extreme Pro 960GB EK D5 PWM Acer X34A / XB321HK 
KeyboardPowerCaseAudio
Logitech G810 EVGA 750 G2 In-Win D-Frame Mini Meridian Explorer 
  hide details  
Reply
Code55
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-6700k 4.8 Asus Impact VIII / EK M8i Monoblock SeaHawk EK X 1080 Ti G.Skill Ripjaw V 2 x 16GB 3733CL15 1T 
Hard DriveHard DriveCoolingMonitor
INTEL 750 400GB U.2 SanDisk Extreme Pro 960GB EK D5 PWM Acer X34A / XB321HK 
KeyboardPowerCaseAudio
Logitech G810 EVGA 750 G2 In-Win D-Frame Mini Meridian Explorer 
  hide details  
Reply
post #17699 of 18250
If I may make a suggestion...if you have to original bios with your card you really should be using that and not any other bios that is listed. When altering bios it is not good to flash a non stock bios. You should only be using them for reference and altering your stock bios to match it. If you are uncomfortable doing so then please post your stock unaltered bios along with the bios you like the most and I'll alter it for you. If you have been using non stock bios then this very will is prob the issue.


Always destroying exergy
post #17700 of 18250
Quote:
Originally Posted by romanlegion13th View Post

Nice i seen these did not realize you could hook a GPU up to it..
I think i will wait for the Next Titan then upgrade my whole system to these EK Coolers so easy to install..
no messy custom water Loop

My Hybrid and H100 will have to do untill next year dont want to spend any more on my T-X with water block

I've disconnected my GPU block a dozen times, not even a drop of water. biggrin.gif
    
CPUMotherboardGraphicsRAM
Intel Core i7 5960X Asus Sabertooth X99 Gigabyte 1080 Ti Founders Edition Corsair LPX 
Hard DriveHard DriveCoolingOS
1.2TB Intel 750 PCI-E SSD 2TB Western Digital Black EKWB Predator 360 With One Titan X In Loop. Windows 10 Anniversary Edition 
MonitorKeyboardPowerCase
Acer XB280HK 4K Gsync QNIX QX2710 LED 27" WQHD 2560x1440 Corsair AX1500i Thermaltake Core X9 
MouseMouse Pad
MadKatz R.A.T. 8 Zowie 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i7 5960X Asus Sabertooth X99 Gigabyte 1080 Ti Founders Edition Corsair LPX 
Hard DriveHard DriveCoolingOS
1.2TB Intel 750 PCI-E SSD 2TB Western Digital Black EKWB Predator 360 With One Titan X In Loop. Windows 10 Anniversary Edition 
MonitorKeyboardPowerCase
Acer XB280HK 4K Gsync QNIX QX2710 LED 27" WQHD 2560x1440 Corsair AX1500i Thermaltake Core X9 
MouseMouse Pad
MadKatz R.A.T. 8 Zowie 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] nVIDIA GeForce GTX TITAN X Owners Club