NEW Benchmark comparison between AMD & Nvidia

Discussion in 'General Discussion' started by John, May 2, 2014.

  1. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,083
    Likes Received:
    111
    Using the original idea for a loose comparison benchmark in this thread : http://isiforums.net/f/showthread.php/16951-R9-290-or-GTX-770-4GB

    And using this image as our comparison betwee what was, and what is : by Nuno Lourenço

    [​IMG]

    Download this replay : http://www.mediafire.com/download/iiawknhifwlwb1a/rF2_unofficial_benchmark.7z

    What is needed:
    -rFactor 2 Build:660, Honda Civic and Silverstone 1.14 track
    -Fraps->http://www.fraps.com/


    How to test:
    -unzip the bench.Vcr file to "rFactor2\Userdata\Replays" folder
    -use the rFactor2 settings shown in screenshots
    -load the replay
    -press ENTER (for fullscreen)
    -press INSERT x 3 (cockpit view)
    -start FRAPS benchmark when buzzer sounds and lights go out
    -stop FRAPS benchmark after two laps on the finish line.

    My Results :

    GPU = Nvidia GTX 780 Ti @ 875mhz (stock)
    CPU = Intel I7-950 @ 3.07ghz (stock)
    Driver = 335.23

    2014-05-02 17:30:10 - rFactor2
    Frames: 10786 - Time: 168419ms - Avg: 64.043 - Min: 51 - Max: 87
     
  2. MarcG

    MarcG Member

    Joined:
    Oct 4, 2010
    Messages:
    6,407
    Likes Received:
    1,705
    Should'nt we all do this with a default PLR as well? pretty sure changes to that happen in Replays (therefore rendering different results) but not 100% sure so could be wrong :confused:

    Specs on left under Avatar:

    2014-05-02 10:24:51 - rFactor2
    Frames: 5112 - Time: 156062ms - Avg: 32.756 - Min: 9 - Max: 61
     
  3. Murtaya

    Murtaya Registered

    Joined:
    Jan 25, 2012
    Messages:
    531
    Likes Received:
    0
    What in game settings? Haven't they changed since the first time this benchmark was used... soft particles etc? Sorry just saw at the top of the screenie, max settings, not sure that 3x AA actually works in all cases though. Should really be 8AA. If you want to test your machine you try to cripple it, as Marc says, the clamp setting may or may not make a difference as well.
     
    Last edited by a moderator: May 2, 2014
  4. MarcG

    MarcG Member

    Joined:
    Oct 4, 2010
    Messages:
    6,407
    Likes Received:
    1,705
    it's in the zip file....
     
  5. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    577
    Likes Received:
    26
    Its all in .7z file.

    My Results:

    GPU = Nvidia GTX 780 Ti
    CPU = Intel I7 3770K @ 4.5ghz
    RAM = 4x4GB DDR3 2133mhz
    HDD = OCZ Agility 3 120GB
    Driver = 337.50

    1920x1080 (GPU default clocks)

    Frames, Time (ms), Min, Max, Avg
    12422, 156435, 64, 103, 79.407
    _________________________________________

    1920x1080 (GPU + 140mhz core / +340mhz mems)

    Frames, Time (ms), Min, Max, Avg
    14106, 156980, 71, 126, 89.859

    @ Jonh, try to overclock your CPU at least to 4ghz to give some more freedom to your 780Ti. You must be bootlenecked too with PCI-E 2.0 slot right?
     
  6. Tosch

    Tosch Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,872
    Likes Received:
    51
    My Results:

    GPU = Nvidia GTX 670
    CPU = Intel i5 2500K @ 4.4ghz
    RAM = 4x4GB DDR3 1333mhz
    HDD = Crucial M4 120GB
    Driver = 337.50

    1920x1080 (GPU default clocks)
    Frames, Time (ms), Min, Max, Avg
    7823, 156562, 38, 71, 49.967


    1920x1080 (GPU +90mhz core / +300mhz mem)
    Frames, Time (ms), Min, Max, Avg
    8052, 156874, 39, 75, 51.328
     
  7. deak1944

    deak1944 Guest

    My Results:

    GPU = Nvidia GTX 570 2.5GB
    CPU = AMD FX-8350 @ 4.2ghz
    RAM = 4x4GB DDR3 1866mhz
    Driver = 335.23
    Frames: 7973 - Time: 156594ms - Avg: 50.915 - Min: 38 - Max: 66
    Default player.ini

    GPU temp = 78c 99% usage
    2164MB GPU mem
    page file 6016 MB
    rFactor 2 Working Set memory 3193K 3D settings: Max pre-rendered frames 1, Multi-display= Single Display, Power management = Max

    Test 2 same as above but with default Nvidia settings
    Frames: 7566 - Time: 156078ms - Avg: 48.476 - Min: 38 - Max: 63
     
    Last edited by a moderator: May 3, 2014
  8. G.Bird

    G.Bird Registered

    Joined:
    Apr 20, 2013
    Messages:
    52
    Likes Received:
    0
    OS:Windows 7
    Home Built
    Mainboard:ASUS
    CPU:Intel i5 760 @2.8GHz
    RAM:8Gb
    Graphics:ASUS 660ti 2GB

    Result.
    rFactor2
    Frames: 9070 - Time: 152771ms - Avg: 59.370 - Min: 48 - Max: 78
     
  9. madquad

    madquad Registered

    Joined:
    Jan 31, 2014
    Messages:
    34
    Likes Received:
    11
    OS: Windows 8.1 64
    Home Built
    Mainboard: ASUS
    CPU: Intel i7 3770k
    RAM: 8 GB
    GPU: HD7970 gh ed. Result. rFactor2 Frames: 9150 - Time: 156625ms - Avg: 58.420 - Min: 47 - Max: 76
     
    Last edited by a moderator: May 2, 2014
  10. Nuno Lourenço

    Nuno Lourenço Registered

    Joined:
    Oct 19, 2010
    Messages:
    577
    Likes Received:
    26
    We have some strange results here... I think that some guys are not using default driver settings or correct game settings and that make comparision unfair...

    I see a GTX 570 with same fps than GTX 670... then a 660Ti 20% faster than a GTX 670.....

    Strange....
     
  11. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    42
    Strange do tell :)


    I championed ingame benchtests for sims for 10 years everyone said negative or not nice things to me.


    Oh well


    Is that with or without this new AMD driver. ( without I am guessing )

    Either way pretty much sucks for ATi from what I looking at.


    Yet in another thread I get pillared for saying I would have to be silly to buy a AMD.

    Well sorry looking at that you would be..........okay.




    edit: So lets see that 100% AMD FPS improvement please ( sniggers )
     
    Last edited by a moderator: May 2, 2014
  12. Murtaya

    Murtaya Registered

    Joined:
    Jan 25, 2012
    Messages:
    531
    Likes Received:
    0
    Back when I had a 7970 and did this test I was only getting 25 fps and now they are getting 47 fps. That's not a bad improvement at all. It looks like they are 100% better than 1 year ago but thats been a gradual ISI related thing, rather than the AMD drivers I think.

    Admittedly it was loch drummond before so maybe not a good comparison. Maybe I need to drink less and read threads better before posting sh!te.
     
    Last edited by a moderator: May 3, 2014
  13. SMOK3Y

    SMOK3Y Registered

    Joined:
    Jun 23, 2012
    Messages:
    80
    Likes Received:
    0
    Triple screen, HDR, AAx4 rest in game Max, NO Multi, NVidia Titan (rest of PC under name)

    2014-05-03 10:03:09 - rFactor2
    Frames: 11619 - Time: 162943ms - Avg: 71.307 - Min: 50 - Max: 93
     
  14. Murtaya

    Murtaya Registered

    Joined:
    Jan 25, 2012
    Messages:
    531
    Likes Received:
    0
    I'm getting a higher average with everything maxed out at 5995x1080 with 8xAA than 1920x1080 with 4xAA? not much in it only 58 against 54 fps but, does that mean my cpu is bottlenecking the 780ti? Surprised if that's what it is.

    Need to retest tomorrow more sober I think.
     
  15. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,083
    Likes Received:
    111

    I did think about using LD again, but it still would not be a direct comparison because LD was optimised after. So I instead went with the content that is included in the current Demo build. That way anyone still looking and deciding can have something they can directly compare themselves. The general idea is to give a current comparison between card make and generation. The previous bench image is a comparison only in regard to the general position on the chart. Hope i am explaining clearly, sometimes I confuse myself :p
     
  16. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    28
    The only way to make sure this is completely fair is to make sure AMD and NVIDIA users each have identical driver settings (default AMD and NVIDIA driver control panels, Radeon Pro, NVIDIA Inspector, etc.) and also the PLR file.
    Is your 780 Ti really running @ only 875 MHz while actually in-game?
     
  17. John

    John Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,083
    Likes Received:
    111
    Mine is a stock non OC card, has a boost clock of 928mhz. I have a number of performance limitations in my PC still, the older model i7-950, the X58 chipset which comes with the older generation PCIe x16 2.0 slot etc etc. I am happy enough with my current performance, the settings in the bench are pretty much how I run it anyway, apart from the resolution. I use a Dell 27" IPS monitor, the UD2711, and always run native res of 2560x1440.

    When you look at the old chart in the first post, and realise my old card was the HD6990 OC that sits 4th last, then compare the exact same PC with the GTX780Ti that sits 5th from the top in that same chart, why would I even worry about trying to tweak the last drop of FPS. I would have to overclock, which on my PC means spending more money for water cooling (at minimum a 240 radiator) or a big air cooler (I hate the thought of that heavy copper block pulling on the MB and all the flex that comes with it).

    I'd rather save my pennies now and build a new platform to use with my GTX card next year. At least then I won't have spent money chasing performance that may not actually be there all things considered.

    So, the crux of it is, I haven't bothered to alt & tab out to look at the voltage as I am really just satisfied with the performance I am getting.
     
  18. mclaren777

    mclaren777 Registered

    Joined:
    Oct 5, 2010
    Messages:
    546
    Likes Received:
    14
    Does the benchmark use Silverstone or Loch Drummond?
     
  19. G.Bird

    G.Bird Registered

    Joined:
    Apr 20, 2013
    Messages:
    52
    Likes Received:
    0
    I can assure you the test was using the settings requested by the Op no point to lie.
    For the test the NVidia control panel was all stock settings nothing changed from default
    using NVidia driver Beta v.337.50.
    Theres a lot that comes in to play when doing these benchmarks not just the game and the settings that are used, like is the rig only a gaming rig
    whats the rig running in the background processes ??
     
  20. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    28
    John, even a stock non-oc'd card will, i'm sure, be running higher than 875 MHz while in-game. I could be wrong but it just seems really low to me. I would have expected at least mid 900s.
     

Share This Page