MSI GTX 970 mini review with rFactor 2 performance

Discussion in 'General Discussion' started by DrR1pper, Sep 25, 2014.

  1. adaptable1

    adaptable1 Registered

    Joined:
    Feb 14, 2013
    Messages:
    112
    Likes Received:
    0
    I just tried this and my frame rates dropped to avg 54 FPS...min 39...max 73. Everything maxed. Reflections on high kills the performance. Everything max'd with reflections on low results in nearly twice the frame rate.

    My point is that the older card scales better to 3 displays and high resolutions.
     
  2. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
    ok but did you mention that before?

    It's interesting that your performance drops by half, effectively making your performance the same as non SLI. LesiU achieved 54 fps average on a single GTX 680 (http://isiforums.net/f/showthread.p...een-AMD-Nvidia?p=271668&viewfull=1#post271668).

    Can you test again with and without high reflections and check your msi afterburner graphs to see what the gpu utilisations are during the bench. I would predict either half utilisation per gpu or one gpu at nearly 0 when using high reflections.
     
  3. adaptable1

    adaptable1 Registered

    Joined:
    Feb 14, 2013
    Messages:
    112
    Likes Received:
    0
    Utilization is within 10% of each GPU whether reflections are on high or low (peaking at about 70% and 80%). Really what SLI brings to the table is the ability to maintain a smooth frame rate with higher resolutions. I don't think single monitor performance is improved by SLI with rF2. That said, until the latest NVidia driver and rF2 updates, SLI actually ran slower than a single card.
     
  4. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
    Although i didn't have GTX 680's, i had 2x gtx 460's and they scalled very well (90+%) with everything maxed in rf2 @ 1080 and that was in 2012/13.

    Where both gpu's at 100% with reflections off? It would be more useful to see your graphs actually. You shouldn't be seeing such severe drops in performance. Maybe if you can include the cpu graphs in the same shots. You don't need to do the entire benchmark, perhaps 30 seconds of each from the same starting position will be enough.
     
  5. Sentri

    Sentri Registered

    Joined:
    Jan 11, 2012
    Messages:
    71
    Likes Received:
    0
    My Zotac GTX 970 just come in. Getting coil whine in the menu, capping my fps to 60 fixes it. Uncapped it doubled my fps over a GTX 580. Hopefully the coil whine goes away with use.
     
  6. Marek Lesniak

    Marek Lesniak Car Team Staff Member

    Joined:
    Oct 6, 2010
    Messages:
    1,583
    Likes Received:
    101
    In the menu, you have like 2500fps so some whine is expected. The most important is do you still have it @100-150fps. If not, then it's fine :)
    btw. please do benchmarks like DrR1pper, so we can compare scores with each other.
     
  7. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
  8. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43
    Problem for me I sell my GTX670OC I only got 1/2 the funds to pay for Gigabyte GTX970 G1. loool

    and lets face it, is overkill @1920 unless you into "other stuff"


    I get good gain running 8xAF vs 16xAF ...... looks the same at 140mph. lol
     
  9. Sentri

    Sentri Registered

    Joined:
    Jan 11, 2012
    Messages:
    71
    Likes Received:
    0
    Did a quick bench:
    Avg: 72.179 - Min: 54 - Max: 98

    Yes, coil whine is only when getting 2000 fps in the menus. I wonder if ISI can do a menu fps cap, I remember Blizzard adding one cause super high fps menus was leading to overheated video cards in Star Craft 2.
     
  10. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43

    Coil whine, your power supply ?

    Nah it wont if it is the PSU, why would it at 100% load ?

    PSU can't handle the card.

    Any whine could damage components even though unlikely its possible.

    I don't risk crap like that.

    I certainly never heard any Nvidia card I ever had whining with a good power supply.
     
    Last edited by a moderator: Sep 25, 2014
  11. Sentri

    Sentri Registered

    Joined:
    Jan 11, 2012
    Messages:
    71
    Likes Received:
    0
    Nope video card. The first video card I ever owned to have coil whine.
     
  12. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
    Ok, that's interesting. You definately followed the instructions to the letter? Everything maxed out, 60 FOV, 1920x1080 and level 4 AA?

    Your performance is 9% higher than mine (Avg: 66.224 - Min: 47 - Max: 97).

    Relative to my GTX 770, you GTX 970 was 25% higher which sounds more reasonable but maybe not as much as hoped for still.

    You were also using the clocks that came with your card? (1076 MHz (base) 1216 MHz (boost))

    this one?

    [​IMG]
     
  13. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43
    you 100% its coming from the video card not the PSU.

    I don't think you are dumb either. hehe :)

    Sometimes can be deceiving is all.

    Mind you I do know video card can make squeaks too.


    quote from somewhere........ along these lines I suggest may be the issue.



    Alternatively depending on hardware combo a video card can squeal straining for enough power.
     
    Last edited by a moderator: Sep 25, 2014
  14. Sentri

    Sentri Registered

    Joined:
    Jan 11, 2012
    Messages:
    71
    Likes Received:
    0
    Yes DrR1pper, that's my card. Tested at stock clocks. Re-ran the benchmark: 72.626 - Min: 55 - Max: 97. This time making sure to match the replay readme settings (Some how I missed setting texture detail to full). I do have a 4790K, with turbo boost that's 100 mhz faster then your cpu clock speed and Haswell is a bit faster then Sandy Bridge.

    DurgeDriven, yes it's from the video card. Even a few redditors have posted about it: http://www.reddit.com/r/nvidia/comments/2h8og5/my_gtx_970_from_zotac_is_coil_whining_is_your/ The 900 cards should be using quite a bit less power then a 570 or 580. Those two cards ran with no whine on my enermax 720w power supply.
     
  15. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43
    Last edited by a moderator: Sep 26, 2014
  16. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    31
    I've only read up to post # 20 in this thread, but I just want to add that I personally believe that there is absolutely no way that a top of the line Sandy-Bridge processor like the I5-2500K is a bottleneck in rFactor 2 - not at 4.3 GHz and certainly not at 4.5 GHz.

    I have an Ivy-Bridge CPU (I7-4930K), and even when clocked at only 3.3 GHz (equivalent to the 2500k/2600k @ around 3.5-3.6 GHz) I can still achieve very high framerates in rFactor 2 with my dual 780 Tis. There is just no way that a 2500K at 4.3-4.5 GHz is a bottleneck in rF2.
     
    Last edited by a moderator: Sep 26, 2014
  17. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
    Thanks. But it's not a cpu issue, the cpu utilisation is not remotely maxing out and my GTX 770 is scoring correct results. Something else is at play here on my system. More testing to try and find out what exactly.
     
  18. oHOWEo

    oHOWEo Registered

    Joined:
    Jul 12, 2014
    Messages:
    176
    Likes Received:
    0
    Good to hear, may my 2500k live a long and prosperous life!
     
  19. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43

    Yeah for sure esp if you run single screen screen / GPU.


    In lot of games you will find a 2500K@4.5GHz same fps as a stock i7, others not. p

    Pulls back about the same in synthetic tests.

    Mine does 5GHz on air testing, no way I getting rid of that with a single screen / card setup.


    Having said that I recently built a H97 Pro with a i5-4590 ( same speed as 2500K ) and you could really feel the grunt in desktop over the Z77/2500K even at default clocks vs 4.5GHz.

    I should have done some encoding tests but almost everything booted smoother if not faster and that was with a new Green 1TB, I have a SSD3
    and my 1600MHz C8 vs C9.

    Not a lot mind you but enough to make me take notice. hehhe



    P.S.


    Do a 3DMark if its going to whine, it will whine trust me. :)
     
    Last edited by a moderator: Sep 26, 2014
  20. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    35
    You sure it wasn't just the fresh install?
     

Share This Page