My FPS nearly halves in the FSX demo when I turn the pretty water on (ie. water effects anything above low 2.x) with my A64 / X800XT system. Do newer video cards from the NVidia 7x00 and ATI X1x00 series exhibit such a performance hit with SM 2.0 effects? Gary

Im getting same performance hit with a X1800XT.

Thanks for the feedback OHN. Perhaps it's an ATI card "feature". Anyone with an nVidia card care to comment?Gary

Here are my system specs:P4 3.4Ghz1GB memoryNvidia GeForce 6800I run at 25fps with these settings12XXx10XXx16kTextures maxedTerrain maxedWater maxedAutogen NormalScenery NormalWeather effects maxedSo far it runs well with the old code at 16k color depth and I don't notice too much of a change right now.

I don't notice much of a performance hit no matter where I move the water slider. I have a (nVidia) 7800GT with a 3700+ cpu. I run 1024x768x32.Rhett

I don't notice much performance impact from the water shader effects either. The big framerate killers on my system are autogen and traffic. With those scaled back to scarce and 10% respectively, my framerate hovers around 25-30fps with the framerate limit set to 30fps and most other settings maxed out. Running at 1280x960x32 on an A64 3500+, 2GB DDR, ATI X850XT.I'm pleasantly surprised by the performance of the demo on my system. The new lighting, atmospheric and environmental effects are jaw-dropping and don't appear to come with a significant performance penalty.

Interesting results there SJDickson, as your system is pretty similar to mine. As mentioned, my frame rate nearly halves with water turned up, but I can run scenery detail at full, and autogen one notch below full with neglible impact to frame rates. I haven't played much with AI yet, but I will do so to see if hits me in the performance arena too.I would have thought we would have experienced slow downs in the same areas. :-hmmm Gary

Same thing here - the SM2.0 water is THE performance killer for me with my 7800GTX card, my FPS is halved or more when it's on.

I just ran a quick set of benchmarks, varying the water shader model and autogen/scenery density to see the relative impact of each. Tests were run at 1280x960x32 on an A64 3500+, 2GB DDR, ATI X850XT with the framerate limiter set to unlimited. I paused the game with the Beaver at 800ft above Princess Juliana airport with an external spot view looking toward a densely forested hillside, trying to get a good balance of water and autogen buildings and trees in the frame. This area is one of the most framerate hungry parts of the demo map so the framerates shown are not necessarily representative. V.Sparse Sparse Normal Dense V.Dense E.Dense1.xMid 43 29 29 28 27 272.xMid 41 24 19 18 16 152.xMax 38 24 19 17 16 15Based on these results, it appears that autogen/scenery density has a larger impact than water shader model at the lower density settings and water shader model has a larger impact than autogen/scenery density at higher density settings.For me, the sweet spot is using 2.x shaders for the water and keeping scenery/autogen density at sparse.

To follow-up, I checked to see what impact setting the land and sea traffic sliders to 0 had. I don't have time to run a battery of tests, so this will just be a single data point. Using the same settings as above with all land and sea traffic sliders at 19%, the framerate settled around 25fps with 2.xMax water and sparse autogen/scenery density. After moving the traffic sliders to 0% the framerate rose to around 33fps.

Major performance hit from all "2.x" water settings on my system (see below).Unless the situation improves once the code is tweaked, I'll stick with FS9 until DX10 cards become reasonable affordable.