I'm preparing to do a "mega" benchmark with rF2 and other sims/games. The benchmark will involve all combinations of single GPU, dual GPU (Nvidia SLI), single monitor, triple monitor (Nvidia Surround), 2D and 3D (Nvidia 3D Vision 2). I'm doing tests to see how I should do the benchmark in rF2 (car, track, replay or live mode, A.I., no A.I. for better consistency if testing in live mode, etc.) and I noticed that my FPS are lower in replay mode.... Is this normal? I thought that there would be more calculations going on in live mode, primarily regarding the CPU and physics stuff, and therefore, if anything, live mode would have lower FPS, or that they would both have the same FPS at the most, but for the live-mode to have more fps than the replay?? It just doesn't seem to make sense... System Specs: - GTX 780 Ti @ 1163 MHz core - 1750 MHz memory - I7-4930K @ 4.5 GHz - 16GB RAM (4x4GB - Quad Channel) @ 1866 MHz - 8-9-9-24-2T - Samsung 840 Pro 256 GB SSD - Windows 8.1 - Single GPU, Single Screen, 2D RFactor 2 Config: - HDR - 1920x1080 - No Sync In-game: - Graphics - ALL MAXED - Steering wheel - OFF - Vertical FOV - 25 - Number Of Effects (Audio) - 64 - Speaker Configuration - 5.1 (windows also set to 5.1) (I have a few things changed in the player.JSON file, but nothing that should be drastic) NVidia Inspector - Antialiasing - Mode - OVERRIDE ANY APPLICATION SETTING - Antialiasing - Setting - 16X [16X CSAA (4 COLOR + 12 CV SAMPLES)] - Texture Filtering - Anisotropic Filter Optimization - ON - Texture Filtering - Anisotropic Sample Optimization - ON - Texture Filtering - Quality - HIGH PERFORMANCE - Texture Filtering - Trilinear Optimization - ON - Maximum Pre-Rendered Frames - 1 - Power Management Mode - PREFER MAXIMUM PERFORMANCE Civic, Indy 2007 - 1 full lap alone (no opponents) from pit-out to start/finish line LIVE ----- Avg: 160 - Min: 123 - Max: 185 REPLAY - Avg: 151 - Min: 116 - Max: 172 Civic, Silverstone GT - 1 full lap alone (no opponents) from pit-out to start/finish line LIVE ----- Avg: 162 - Min: 125 - Max: 186 REPLAY - Avg: 153 - Min: 120 - Max: 174 Civic, Silverstone International - 60 second benchmark - begins as soon as "Race" button clicked, I stay in neutral and do not move the whole time) - Steering Wheel - ON - Opponents - 17 - Starting Position - 18 - Race Start Type - STANDING START LIVE (run 1) ------------------------------------------------------Avg: 116 - Min: 86 - Max: 129 LIVE (run 2, "restart race" rather than exiting and reloading) Avg: 119 - Min: 95 - Max: 129 REPLAY ---------------------------------------------------------- Avg: 104 - Min: 69 - Max: 124 Civic, Silverstone International - 2 laps race distance, I stay in last place the whole time but follow the next car closely for the entire race) LIVE --- Avg: 115 - Min: 72 - Max: 147 REPLAY - Avg: 84 - Min: 47 - Max: 106 Quite severe differences in fps between live and replay modes, especially with regards to avg. and min. fps when other opponents are involved (last two tests).
May I have your problems, please? I would suffer quietly, I promise. Replay is not only replay. You can resume while watching it so I think it not only calculates more objects because more things can be seen on the screen while watching everything from TV or other cam but also prepares itself for resuming if there's need for it. That's my guess. But again... Your problems... mine? Yours... mine? Maybe? Perhaps? Please?
It´s not only preparing for resume it will most of the time (as long as cars are on track) have to calculate the real time event too! So it´s more work, not less.
My understanding is the physics have their own thread handled by one core and the graphics have their own thread handled by another core. There is the option of an on-screen display to indicate relative activity in the graphics and physics thread. In general the graphics thread is the heaviest thread so this will limit framerate even when there are no physics. You can for example turn off the new tyre model but while it reduces the cpu power required for the physics thread, it doesn't improve fps. It doesn't explain why your replay framerates are lower, just why you shouldn't expect them to be higher.
You may be getting the speed of the memory and the amount of V.RAM mixed up. 1750 MHz is what the memory on my GPU is clocked at, memory usage would be MB (megabytyes).