vmounier

And just to throw one last option out there, MSI Afterburner has a nice utility that you can overlay or hide with a shortcut, customizable to report on FPS, VAS, GPU, CPU etc. It has become my main tool even though I keep FSUIPC's chime on as a last-minute reminder. Only disadvantage of Afterburner is that it also gives you total VAS used rather than total remaining.

I don't remember how hot my 6700k runs with HT ON and OC 4.6Ghz but not that hot (I'm water-cooled). But yes, definitely more VAS used, so I am back with HT OFF, OC 4.8Ghz and the best performance I've had so far. Of course recently adding a G-Sync monitor help much with the stutters as long as I can stay above 35 fps or so... To be honest, I have not noticed much difference in either FPS or fluidity. I was achieving about the same with HT OFF or ON, and was going to stick with ON so that the computer benefits, but the VAS issue killed the deal. I'm going to try what Steve suggested, just to be sure: to not lasso the sim and apply AM through cfg instead, but I don't expect different results as I'm 99% sure I've tested this before starting to use Lasso. Vince

Good point. I do not use anything to start P3D but I do use Process Lasso to apply AM. However that's very recent and I'm pretty sure I had the same results previously setting AM through .cfg. And even through Process Lasso, I have tested multiple scenarios (HT ON 85, 84, 116 and HT OFF 14, 15) and results are consistent, less VAS used with HT OFF.

Agreed, that would be my assumption too, except that 4 LP's with HT ON (say AM 85 or 116) and 4 cores with HT OFF and no AM still yields a big difference in VAS used in my case. Can't explain it but very real. Any ideas?

I'm a bit surprised no-one has mentioned VAS use. I'd love to go back to HT and AM 116 to take full advantage of my 6700k, but as Rob discovered a while back with his own very different setup, turning HT on has a huge impact on VAS and the extra 500 Mb or so it uses in my case mean an OOM on final after a single pattern at CYVR in the 777.

No obvious improvement by increasing the draw distance. For me the fix is simple to keep the shadows pale and avoid shading them, it makes the effect less apparent.

Same here, and it didn't really bother me that much until I started experimenting with shaders that darken cloud shadows, at which point that small quirk becomes a major annoyance. The problem is not so much lateral screen edges, but it is straight ahead when panning up or down with TrackIR. It's as though the cloud shadow rendering area was actually a rectangle that doesn't cover the entire screen but rather only two-thirds, with the top third left out as well as the far sides. So if looking dead ahead, in the distance there is a well defined horizontal line marking the end of shadow rendering, but if I look up that lines moves up and shadows are rendered much further ahead. Same in the lower corners when panning. It completely destroys immersion, especially when in an outside view. I've tried playing with all the variables available but so far nothing. GTX 980ti, P3D V3.2, but I had that before with a different card and earlier P3D version I'm sure. I actually have not noticed any difference in rendering when changing the cloud shadow cast distance in settings (other than setting it to zero of course), and will have a look at the CASCADES and LOG_PARTITION variables, but not much hope...