rFactor 2 DX11 Benchmark

Discussion in 'General Discussion' started by Nuno Lourenço, Aug 14, 2017.

  1. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    Hello guys.

    We all noticed that rFactor 2 as become visualy better with implementation of DX11 but that cost us all some fps.
    Similar to what I done a couple years ago I created a replay file that we can use as a benchmark and see the real performance that we have now with all our GPU's.

    I already had done some testing in my system that I share here. If you send some result I will update the graphics.

    My system:
    I7 3770K @ 4.4
    16GB DDR3 @ 2133mhz
    GTX 980Ti Gaming 6g
    SSD Samsung 850 Evo 250GB

    Testing each resolution takes only 5 minutes (Loading track and 2 laps race replay)

    If you're willing to compare some results just download the replay file and follow the procedure.
    Thank you guys ;)

    Download replay file: here

    Procedure
    1- Install Fraps (http://www.fraps.com/download.php).
    2- Make sure you select a benchmark key for Fraps. Only nee to save minimu, maximum and average results.
    3- Extract replay file to your rFactor 2 replay folder.
    4- Reset your graphic drivers to default values (otherwise won't be a "fair" comparison).
    5- Make sure you set every graphic option in game as in my screenshots.
    6- AntiAliasing must be set in game options and not in drivers!! For Nvidia cards Max AA level should be 5 (It is in my case), for AMD cards I don't really know what is the maximum in game option. (FXAA is always off!)
    7- Enter the game, go to replay option (make sure you set fidelity to FULL) and load replay. Once in track, CHOOSE COCKPIT CAMERA and press benckmark key when green flag is out! Press it again when you finish the second lap at finish line.
    8- If everything was done correctly you should have a *.csv file in your "C:\Fraps\Benchmarks" folder with something like this

    Frames, Time (ms), Min, Max, Avg
    11481, 221547, 39, 62, 51.822

    Share your results along with your system in forum :)

    Note: For comparison purposes please do the test with all graphic settings as I did otherwise we won't be able to compare it. Thank you
    Note 2: Benchmark as been made with DX11 beta version of rFactor 2 and post effects on ultra.

    [​IMG] [​IMG] [​IMG]

    [​IMG]

    [​IMG]
     
    Last edited: Aug 25, 2017
  2. CamiloNino

    CamiloNino Registered

    Joined:
    Apr 5, 2017
    Messages:
    153
    Likes Received:
    131
  3. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
  4. Emery

    Emery Registered

    Joined:
    Oct 24, 2010
    Messages:
    2,789
    Likes Received:
    1,381
    Nuno is essentially updating an old benchmark we had when rF2 was in its first year or two. It was very useful for helping people choose an appropriate video card, so it is definitely time for an updated version!
     
  5. MarcG

    MarcG Member

    Joined:
    Oct 4, 2010
    Messages:
    6,731
    Likes Received:
    2,078
    Thanks for doing this @Nuno Lourenço (again), one thing though is that Zandvoort is clearly still in an early beta stage, so probably not the best choice of track as further optimisation will likely follow and therefore skew results greatly.
     
  6. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    I did not remember that detail Marc... I'll upload a diferent replay file asap.
     
  7. MarcG

    MarcG Member

    Joined:
    Oct 4, 2010
    Messages:
    6,731
    Likes Received:
    2,078
    No worries mate, Silverstone/Mills/Toban all updated and unlikely to get any major future updates so they would be safe to use amongst a couple of others I think :)
     
  8. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    I used Silverstone and Radical now. Link was updated and my tests too.

    Tests about PCI-E gens I let the older ones (The diference will be similar no matter the track used).

    Feel free to download and test guys.
     
    Last edited: Aug 14, 2017
    MarcG likes this.
  9. Korva7

    Korva7 Registered

    Joined:
    May 22, 2014
    Messages:
    694
    Likes Received:
    200
    Would be interesting to see how those new AMD Vega cards runs rf2.
     
    Chris Lesperance likes this.
  10. Emery

    Emery Registered

    Joined:
    Oct 24, 2010
    Messages:
    2,789
    Likes Received:
    1,381
    GTX 1080 stock, i7-4770 @ 3.4 GHz

    Note I only recorded 1 lap rather than both laps.

    Min Max Avg
    27 57 41.183 -- 5760x1080 multiview on, Level 5 AA
    29 65 45.282 -- 5760x1080 multiview on, no AA
    (multiview on = angled monitors)

    46 75 60.838 -- 5760x1080 multiview off, Level 5 AA
    52 88 70.255 -- 5760x1080 multiview off, no AA
    (multiview off = one extra wide monitor, no angles)

    68 138 98.860 -- 1920x1080, windowed, Level 5 AA
    72 149 105.106 -- 1920x1080, windowed, no AA
     
    Korva7 likes this.
  11. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    Graphics has been updated.

    You are losing fps for not recording the second lap Emery. More importante than that, in 1920x1080 you're losing fps by not having your CPU overclocked :).

    Thank you for your results!
     
  12. Mulero

    Mulero Registered

    Joined:
    Jan 16, 2012
    Messages:
    351
    Likes Received:
    153
    No difference here in rF2 doing OC in a 7600K. Tested a few years ago with an Athlon II X2 2.4Ghz @ 3.6Ghz and no difference either. Nor did I see any difference with a Phenom II 3.4 Ghz @ 4.6 Ghz. With the Athlon I tried doing underclock and performance was only affected when lowering 1.8 Ghz, although this was several years ago already.

    If I have a while this afternoon I will do some OC tests and the benchmark with the 7600K and a RX570 4Gb.
     
  13. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,085
    Likes Received:
    113
    The replay starts in TV cam, I assume you want us to run it using Cockpit cam though ???
    i7 - 4790K @ 4.0 GHz
    16GB DDR3 @ 1600 MHz
    ASUS ROG GeForce GTX 1080 Strix Aura 8GB @ 1695MHz (ie. stock)

    1080P PP Ultra = Avg: 112.753 - Min: 64 - Max: 158

    1080P No post processing = Avg: 134.940 - Min: 70 - Max: 203

    1440P PP Ultra = Avg: 89.466 - Min: 56 - Max: 117
     
  14. Emery

    Emery Registered

    Joined:
    Oct 24, 2010
    Messages:
    2,789
    Likes Received:
    1,381
    Extra lap only increases average and/or maximum fps. More interesting is the minimum fps, because that's what you have to suffer most. I'll do an update with both laps, but it will have to wait until next week.
     
  15. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    Yep John, was to choose cockpit camera. That 1080p results are with or without AA?


    I know that, I tried to do a 1 lap race but minimum is 2... I know that minimum fps is the most important the thing is, if we don't make all the same way, the only results we can compare are the minimum fps.
     
    Emery likes this.
  16. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,085
    Likes Received:
    113
    Level 5 AA in game.

    I normally run at 1440P with everything at maximum settings except post processing off. The extra fps are worth more than the post processing effects to me.
     
  17. Korva7

    Korva7 Registered

    Joined:
    May 22, 2014
    Messages:
    694
    Likes Received:
    200
    GTX 780Ti (stock), i5-4670K @ 4,2 Ghz

    Min Max Avg
    22, 63, 39 -- 1920x1080, windowed, Level 5 AA
    23, 75, 44 -- 1920x1080, windowed, Level 0 AA
     
    Last edited: Aug 16, 2017
  18. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    592
    Likes Received:
    63
    Graphics updated!

    Thank you!
     
  19. patchedupdemon

    patchedupdemon Registered

    Joined:
    Jan 3, 2017
    Messages:
    1,602
    Likes Received:
    1,151
    Are the physics being run/calculated as they would during a live lap/race.

    A similar thread has just created in iracing,yet the general consensus is using replays isn't a true representative of performance because the physics thread isn't no where near as active in replay vs live.

    Hence for the need of a ingame benchmarking too provided by the devs
     
  20. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,085
    Likes Received:
    113
    Yes, but even if it didn't, the graphics performance would still be measurable between different GPU's.
     

Share This Page