Released NEW! Endurance Pack - Now Available!

Discussion in 'News & Notifications' started by Christopher Elliott, Jun 15, 2018.

  1. Filip

    Filip Registered

    Joined:
    Feb 9, 2017
    Messages:
    1,208
    Likes Received:
    937
    What???
    One user is having 5-7% improvement, other is having 80% improvement :rolleyes: and most don't have any improvements at all.
     
    UsedMomo likes this.
  2. vittorio

    vittorio Registered

    Joined:
    Jan 11, 2012
    Messages:
    1,118
    Likes Received:
    540
    Thats why I think it is important for people to post performance measurements! At the moment only MarcG does that. And MarcG having ~80% speedup. So post your benchmarks, the circumstances and how you measured.
     
    Last edited: Jul 10, 2018
  3. C. Banger

    C. Banger Registered

    Joined:
    Jan 5, 2017
    Messages:
    99
    Likes Received:
    72
    I'm a 'common user' with a low/mid spec PC (i7-4470 + GTX 970). I'm getting approx 10-20% improvement on single monitor so I'm seeing benefits. I also have VR but don't even bother with rain in it.

    Is it a fact that "most don't have any improvements at all"? Where do you get that data from?
     
  4. vittorio

    vittorio Registered

    Joined:
    Jan 11, 2012
    Messages:
    1,118
    Likes Received:
    540
    Regarding measuring performance improvements: This winter we got a performance update. But I didn't see any performance improvements. What was wrong you may ask? I'm only racing in VR. I don't have a performance meter like FPS. With the performance update my computer still struggled to get near 90 FPS with current settings, so why change settings.

    So. What went wrong and still is? I didn't measure performance with something like a FPS meter, but instead only measured with my eyes. Is it above 90Hz? -> no stuttering. Is it below 90Hz? -> stuttering

    The point is: With before software I could have been barely above 45 Hz. With new software barely below 90 Hz. There could have been theoreticaly a speedup of nearly 100% and I wouldn't even have noticed that with the method I measured performance.
     
    Last edited: Jul 10, 2018
  5. marvelharvey

    marvelharvey Registered

    Joined:
    Apr 25, 2013
    Messages:
    410
    Likes Received:
    283
    Optimisation is all about the low hanging fruit. S397 may have looked at the code, seen easy gains to make in wet-VR and doubled the framerate in X number of hours. Whereas wet-2D may already be reasonably optimised and those X hours only given a minor boost.
     
    Jeferson Richart likes this.
  6. davehenrie

    davehenrie Registered

    Joined:
    Jul 6, 2016
    Messages:
    7,454
    Likes Received:
    4,370
    I'm not denying the rain slowdown, but remember, this VR upgrade came AFTER a general DX11 speed-up for almost all users. Only then did they turn their attention to VR users. So this is not like ONLY VR users got an upgrade. They had to wait even longer. Would I like to see more? Yes, but if there are users here...and there are....that are posting 200+ fps on a single screen with an AMD card, then I think there is only so much S397 can do.
     
    Atsuki likes this.
  7. Christopher Elliott

    Christopher Elliott Registered

    Joined:
    Jul 31, 2014
    Messages:
    4,551
    Likes Received:
    7,538
    Of course not everyone will have the same results, and as we said already the rain performance improvement was for VR mainly. Some small improvement may be seen in non-VR as well.
    We also can't account for every single setup out there - once we have a good baseline from testers PCs, and we don't find any new bugs introduced, we release improvements for you as soon as we can.
     
    yippssy, Dookie, RaceNut and 8 others like this.
  8. DrivingFast

    DrivingFast Registered

    Joined:
    Mar 1, 2018
    Messages:
    1,638
    Likes Received:
    1,083
    I am in holidays !
    Hi.

    I am in holidays, so no complete tests.

    I can not give you my numbers of the last build.

    According to my * Impression *, and the minimum FPS numbers I've seen, there is no (or very little) improvement.

    I did a lot of factual and numerical tests before but not specifically on the rain, and anyway on "flying" sheets of paper.

    As for the rain, I did a really fast test in the previous build in ULTRA \ full HD, and my result was :

    -48% FPS in the rain

    mostly cloudy + Atlanta MP + 7h00am until 9h00am + time scale X3 + NO AI
    VS
    mostly cloudy + rain + time scale X3 + 7h30am until 9h00am + NO AI

    I do not know if I activated the mirrors because they DESTROY my FPS and I think the reduction of FPS is more important, I have not fully followed my usual test process.
    I have not used AI, I think it would be much worse with (splashing water, raindrops on cars etc ...).
     
  9. AMillward

    AMillward Registered

    Joined:
    Apr 4, 2017
    Messages:
    1,879
    Likes Received:
    1,837
    I've done mod reviews in the rain and seen no FPS drop. Also ran in heavy spray with no FPS drop.

    Need to test with AI but I don't understand why my average system can run at 60 in the rain while others are dropping.
     
  10. hitm4k3r

    hitm4k3r Registered

    Joined:
    Jun 15, 2016
    Messages:
    1,320
    Likes Received:
    3,121
    Those numbers don't say much if you don't note your hardware configuration and graphic settings, let alone your FPS in dry conditions. If you are getting 150 FPS in dry and 100 FPS with high graphic settings in wet conditions I wouldn't be worried at all. I am not arguing that those numbers from your tests aren't true and there can certainly be people who struggle more with certain hardware combinations even if their setup looks better on paper as even just a hardware driver can cause issues. I am particular interested in those Full HD tests, your hardware and how you adjusted your graphic settings according to the recommendations by S397. Ultra HD will require stronger hardware ofcourse if you want to achieve the same visual quality and performance as with Full HD.
     
  11. Seven Smiles

    Seven Smiles Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,099
    Likes Received:
    1,152
    Ooh! I’m a member of an elite! :cool:
     
    C. Banger likes this.
  12. Seven Smiles

    Seven Smiles Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,099
    Likes Received:
    1,152
    No one seems to have considered that this might have been something that was simple to fix with no downside so S397 just went ahead and did it. The impact of a bug is not related to how hard it is is to fix.
     
    Nimugp likes this.
  13. DrivingFast

    DrivingFast Registered

    Joined:
    Mar 1, 2018
    Messages:
    1,638
    Likes Received:
    1,083
    No CPU real-time issue since the performance update for 50 AI.

    I can also run 80/90 (lol) AI on the 21km Nordschliefe when I create a LAN race with the same computer with a low power full i5 2500K not over clocked. But the GPU gtx 780 suffer even in dry/standard conditions.
     
  14. hitm4k3r

    hitm4k3r Registered

    Joined:
    Jun 15, 2016
    Messages:
    1,320
    Likes Received:
    3,121
    That's pretty wierd then. Because I am running gtx 770 and it works ok. Not in full HD, but when you tweak the graphic settings according to your hardware it should run just fine. Keep in mind that those gpu's are half a decade old by now so we shouldn't expect miracels if we are realsitic.
     
  15. Marcel Offermans

    Marcel Offermans Registered

    Joined:
    Oct 4, 2010
    Messages:
    647
    Likes Received:
    2,936
    If this were a simple issue, it would have been part of the first round of performance improvements already. It was hard to fix, and as @Christopher Elliott said earlier here, it was our specific goal to fix rain in VR. According to overwhelming feedback we got, that has worked. Some are even reporting a slight improvement in other scenarios. That's a plus.
     
  16. bbspadon

    bbspadon Registered

    Joined:
    Jan 15, 2017
    Messages:
    53
    Likes Received:
    252
  17. vittorio

    vittorio Registered

    Joined:
    Jan 11, 2012
    Messages:
    1,118
    Likes Received:
    540
    Thanks for your tests and results!

    You say you have -48% FPS in rain and otherwise same conditions (when I understood right with not the latest rain optimized build).
    1. I would be very interested about your absolute FPS (not only relative in %).
    2. You wrote you did your tests using Road Atlanta track which is IMHO not an official S397 track. Unofficial content possibly is very unoptimized and not telling anything. Could you do your tests again with official S397 content please, track and car?
    3. You were speaking about Ultra HD and Full HD. Were these tests done in Ultra HD or Full HD? I can imagine that Ultra HD is a problem. I'm not an expert but the GFX card fill rate then is a big factor (GTX780 is already old too) and in combination with rain possibly a lot more (transparent objects all over the screen).
    4. whats your absolute FPS (dry/rain) with latest optimized build.

    I noticed you have an i5 2500k@stockGHz/GTX780. I have an Intel i7 3820@4.5GHz/GTX780Ti, which is similar. When I simraced with that hardware in single screen, well already few years ago, I never had performance issues. Now I only race in VR, which is a totally different beast (like half FPS but you need >90FPS in VR), and I have no new single screen data anymore. Using single screen I was always above 100FPS with near max settings with that hardware. Performance wasn't an issue anymore, I didn't spent time to optimize anymore. It was a different story with my former hardware AMD Phenom II X4 965/GTX570. With that hardware I was constantly optimizing and doing performance tests (min 30 FPS at some performance tests) using single screen.

    I did my single screen performance tests few years ago. So I was using DX9. FPS always was more then 100 FPS. I remember MarcG did DX9 vs DX11 tests, and latest DX11 optimizations were on par to DX9 FPS. So the switch DX9 -> DX11 shouldn't be an issue anymore either.
     
    Last edited: Jul 11, 2018
  18. mesfigas

    mesfigas Registered

    Joined:
    Oct 15, 2015
    Messages:
    1,722
    Likes Received:
    830
    i think VR users need minimum Nvidia GTX 1060 3giga or RX580 4 giga
    at least this is what i understood when i was searching for VR info
    all VR users agree that good VR experience comes from 1080 GTX but still needs more power that that
    VR generally in all games is on a very Premature state
    virtual technology is very expensive and everybody knows that people who " own latest tech" they sell it with the little spoon if you can dig what i say
     
    AMillward likes this.
  19. Ari Antero

    Ari Antero Registered

    Joined:
    Jul 27, 2012
    Messages:
    1,882
    Likes Received:
    829
  20. vittorio

    vittorio Registered

    Joined:
    Jan 11, 2012
    Messages:
    1,118
    Likes Received:
    540
    I wouldn't say VR is premature. But, yes you are right. It is as it is, VR in it's current state is very performance demanding (without eye tracking and foveated rendering). And it is no surprise either, cause in the end rendering 2 screens in ~FullHD.

    But then, I have zero problem racing in VR in iRacing and Assetto Corsa. But then I know too - that physics, tyre, FFB calculation is a lot more intensive in rF2. Just remember tyre physics is calculated in 2400Hz. And because of that quality rF2 will always be my simulation (till a more realistic sim is available).

    I know my GTX780Ti is old (and I want to wait for next Nvidia Ti series). So, to race in VR in rF2 I have to make some restrictions . No multiplayer, no AI, no mirrors, minimal settings... But thats OK for me. But also because I only race at AC Nordschleife convertion which is not optimized, S397 content probably is OK and a lot more optimized. -> we need official laser scanned Nordschleife by S397 or partners :)
     
    Last edited: Jul 11, 2018

Share This Page