This is sort of a spin-off from my other thread but i'm hoping to get a real answer on this issue from someone who knows and I don't think it'll happen in the other thread since it's buried in mix of things. So about a year ago, Tosh told me to use max pre-rendered = 1 which resolved the issue of my cpu from being at 100% utilisation down to no more than 15-25% on each core (i5-2500k @ 4.5ghz) whilst in rf2 (i think with many AI and/or in a replay benchmark). I recently tested this again and with 2 different graphics cards (to make sure it's not the gpu) and the same still holds true today. In terms of performance, i found a 16% fps loss and a stutter without max pre-rendered = 1 when performing a replay benchmark at max settings. The fps loss is explainable by the bottlenecking of the cpu (essentially) but the real question is why I must manually set max pre-rendered = 1 for rfactor 2 in order to avoid this cpu bottlenecking resulting in a 16% fps loss. Since max pre-rendered = 1 is not a default setting in the nvidia control panel, i therefore wonder how many other rfactor 2 players are unaware of this and suffering a similar and completely unnecessary performance drop. edit: It would seem not everyone needs to use pre-rendered frame = 1. Also, that performance loss without max pre-rendered frame = 1 is now happening happening regardless of it on or off (but it does drastically lower cpu utilisation from spiking to 100 down to no more than 25%.) Something else is causing rf2 to act up on my system having tested with 2 different graphics cards and the same result. So please take this post with a pinch of salt.
It is a default setting in my NvCP. It's set to 1 as default (344.11 drivers). The name has been changed to "Virtual Reality Pre-Rendered Frames".
I'm always using Vsync and with another value than 1 in the max pre-rendered I got very light stutters. Only the value of 1 gives me a real smooth experience.
and i found this, posted sometime ago: note: Replay of the same benchmark recording immediately after 2 laps with identical graphical settings. post: http://isiforums.net/f/showthread.p...-my-gpu-please?p=225994&viewfull=1#post225994
Use Nvidia Inspector http://www.guru3d.com/files-details/nvidia-inspector-download.html View attachment 14360
There's no such a setting like "no max prerendered frames". Higher CPU usage is because you have more than 1 prerendered frame (which require more work from CPU).
It gives you the lowest lag possible, because only current frame will be rendered to frame buffer and then sent to monitor.
Seems like it's just for rF2 but should make some more tests. rF2 is the only game I really play, I usually quit the other ones after 5 minutes...
Sparse Grid Super Sample and other features http://www.overclock.net/t/1250100/nvidia-sparse-grid-supersampling
And after you have seen rF2 with SGSS, you will have a hard time turning it off and going back to any other setting...even though it is fairly resource heavy.
This setting is surely a 'game-changer'. I just tried it. Getting between 67-75 fps with a GTX670 all things turned way up in game...except texture filtering (at 4x) movies, wind/crowd off and 24 cars visible. Could be placebo but the steering wheel inputs feel more pointed as well? The whole game feels better. Super nice, thanks.
the sim gets easiest to drive around 100 fps - I don't think it gets any better at greater than 100 fps (at least for me) - its not placebo - its this way in all sims for me.
There actually few things witch is greater then 100 fps Locked 120 fps = 0 input lag. Lisghtboost = 0 motion blur. Nvidia 3D Vision = All game options max, HDR on, triple screens, Multiview on, 8X SGSS , 16XSupersampling = 60+60= 120 fps
Sorry for the OT, but in the directions DD posted about Sparse Grid Super Sample, it says this: View attachment 14366 Anyone know the AA compatibility bit for rF2?