Just checked, and it's happening on my system too.

 

I'm running P3D 2.5.12945 on Windows 10 Pro Insider Preview build 10074 (slow update setting) with a Titan Black on driver 352.84 (latest driver for Windows 10).

 

The PCIe Link State setting is off and the Nvidia driver is set to "Prefer maximum performance".

 

Not sure what impact it has, but worth looking into.

(Edit -> Oops missed kmagelie reply which puts my theory of it being GPU architecture into doubt!).

 

 

So it looks as if a pattern maybe emerging based on GPU (at least with the limited sample population so far):

 

Dave_YVR : 970GTX - Win7 64

DylanM : 980GTX - Win 8.1

shaddyy : 980GTX - Win 8.1

myself : 980GTX - Win 8.1 / Win 10

 

Although if this is the case then it looks limited to the GM204 architecture as Jay is not seeing the anomaly with his Ti (that's with a different driver which I'll try after work).

Running Windows 8.1, GTX980, driver 353.06, P3D with water @ Ultra and my profile stays and never moves from P0.

 

Flying right now and monitoring.

Post by Beau Hollis (Prepar3D Rendering System Lead) over at LM forums:

 

http://prepar3d.com/forum-5/topic/selecting-ultra-water-detail-setting-causes-p2-state-lock-in-on-nvidia-drivers/#post-109054

 

 

----------------------------------------------------------------------------------------------------------

Quote:

When Ultra water is enabled, the 3D waves are enabled. The 3D wave simulation relies on CUDA when using NVidia hardware. My guess would be that either:

a) In the case of using CUDA and Rendering pipelines in the same app the driver drops the level to reserve resources for CUDA

b) The level drop is a because our water simulation does not build against the most current cuda libraries.

 

When last I tested, the cuda based wave simulation was faster than the Direct Compute version (which is used by any cards lacking cuda support). If you want to try using Direct Compute, you can do the following:

1) Open File: Prepar3D v2\TritonResources\Triton.cfg

2) change “disable-cuda = no” to “disable-cuda = yes”. This should be around line 128.

3) Try to fight the urge to tinker with the rest of the config options.

4) If you fail at step 3, please back up your config file first. It’s very easy to break the water or even cause crashes by changing settings in this file because we only use a small subset of the Triton library and we modified the source to disable features we’re not using.

 

----------------------------------------------------------------------------------------------------------

3) Try to fight the urge to tinker with the rest of the config options.

4) If you fail at step 3, please back up your config file first. It’s very easy to break the water or even cause crashes by changing settings in this file because we only use a small subset of the Triton library and we modified the source to disable features we’re not using.

 

Love the support from the Devs...and it's obvious they know their user base quite well by now

 I just tried the suggested fix mentioned above and can confirm that it does now set the P state to 0 with water on Ultra.

 

 Cheers to boez for linking the fix and of course to Beau for the support.

Not tried it myself yet but glad to hear Dave had success.

 

Curious as to what is the cause on certain PCs that are affected. One thing I remembered is that when I install nVidia drivers I always deselect everything except the graphics driver...could that be a factor?

Confirmed that Power State now remains at P0.

 

I've posted on LM thread asking how the performance of non-CUDA ultra water detail (i.e. the fix) is compared to the default situation of using CUDA. Even though I can now select that setting I would like to think the hit I'm getting is not too big (if at all).

GPU load, thermal, and voltage are well below their thresholds with Ultra enabled,

 

Certain nVidia cards drop down their clock a notch at 70°F, but only using specific drivers. For example, take a look at how  my old GTX 680 handled temperature.

 

EDIT: I should have added a link to the article, in case anyone wanted to test their GTX 980 with GPU-Z:

 

http://www.overclock.net/t/1265110/the-gtx-670-overclocking-master-guide

 

The guide is not for our generation cards, but the procedure for testing the boost clocks is pretty much the same. I would also suggest perusing the nVidia message boards, because there are a few threads about the thermal behavior of Maxwell cards with recent drivers being a bit squirrely. I didn't post any links, because the threads seemed more speculative than helpful.

Jay

 

Interesting graph but

I can turn every other setting completely to the left and yet still see P2 if water at ultra. I'm pretty certain it's not thermal protection kicking in. My open frame case means my Gpu temperature barely goes above 60degC even during high loads. (Scottish weather helps too!)

 

LM have provided a work around based on removing CUDA from the ultra water calculations although this doesn't explain the cause but does remove the effect with all else unchanged including average temperature.

 

Andy

Did you notice any change in performance? More/Less framerate, More/Less smoothness?

Jay

 

Interesting graph but

I can turn every other setting completely to the left and yet still see P2 if water at ultra. I'm pretty certain it's not thermal protection kicking in. My open frame case means my Gpu temperature barely goes above 60degC even during high loads. (Scottish weather helps too!)

 

LM have provided a work around based on removing CUDA from the ultra water calculations although this doesn't explain the cause but does remove the effect with all else unchanged including average temperature.

 

Andy

Andy,

 

I'm not a guru regarding NI, but can't one set the clock speeds with NI for the P2 state? Maybe you can use NI to "trick" the driver into behaving "normally". Maybe someone who knows more about nVidia Inspector can chime in. See for example, this somewhat unrelated thread:

 

http://einsteinathome.org/forum_thread.php?id=11044

 

 

Jay

You may look here. http://forums.guru3d.com/showthread.php?t=399677

Spirit

Thanks Spirit - downloaded and tried it with the 353.12 drivers but still get the P2 state.

 

Did you notice any change in performance? More/Less framerate, More/Less smoothness?

 

Can confirm there is a significant difference in framerate by switching the flag - this is likely because of the switch back to P0 state (as I can't do an apples to apples comparison between Direct Compute and CUDA with the bug in play).

Andy,

 

I'm not a guru regarding NI, but can't one set the clock speeds with NI for the P2 state? Maybe you can use NI to "trick" the driver into behaving "normally". Maybe someone who knows more about nVidia Inspector can chime in. See for example, this somewhat unrelated thread:

 

http://einsteinathome.org/forum_thread.php?id=11044

 

 

Jay

Hi Jay

 

Yes it is possible to do this and I spent time running like that before I thought 'why should I need to do this?' and hence started this thread.

 

You simply select the P2 drop down box in the NI overclock page and change as desired. By doing so you can get as high as P0 default values. BUT I've no idea whether P2 slows down any other part of the GPU other than clock speed and so a answer from LM was preferable.

 

A.

I think Beau answered your question as best he could. I vote for the option that P3d is not compiled with the latest version of CUDA.