The GTX 780 TI SLI - End of Life - Driver Performance Analysis

Discussion in 'Videocards - NVIDIA GeForce' started by Cain, Jun 28, 2015.

  1. Cain

    Cain Member

    Messages:
    27
    Likes Received:
    0
    GPU:
    nVidia GTX 1080 SLI
    The GTX 780 TI & SLI - End of Life - Driver Performance Analysis

    The GTX 780 TI & SLI - End of Life - Performance Analysis

    The point of this post / thread was to display some hard facts and anecdotal experiences with the GTX 780 Ti SLI.

    Kepler Degradation In Performance:
    For something that could be tested, it seems like a lot of people have not taken the time, an arduous and tedious amount of time, to produce some tangible data regarding Kepler degradation in performance. Popular opinion even states that the last viable driver was 347.88. Well I am here to tell that I have data that may prove otherwise.

    I am going to let you guys make your own conclusion, I have no interest in persuading opinion towards nVidia or even against nVidia. There are somethings that I feel should be corrected, because when people have all the facts, they can make informed opinions themselves.

    My motivation is just help inform users and I feel obliged to do the testing since I own the two cards.

    The list of drivers I have tested:
    • 331.93
    • 332.21
    • 334.89
    • 335.23
    • 337.88
    • 340.52
    • 341.44
    • 344.75
    • 345.20
    • 347.09
    • 347.25
    • 347.52
    • 347.88
    • 350.12
    • 352.86
    • 353.06
    • 353.30
    • 353.38

    Benchmark:
    • 3DMark Fire Strike
    • 3DMark Fire Strike Extreme

    System Setup:
    [​IMG]

    Why the one benchmark? Well initially I had planned to do a few benchmarks. The difficulty arises in my mind when trying to find a benchmark that's been around a while and supported in SLI by multiple drivers. For example GTAV is only good for 3 previous drivers and that wouldn't show enough data in my mind. 3DMark is also something that sells GPUs. This is the de facto standard for reviewing cards/benchmarking cards. In my mind, this is where nVidia, would at least focus support every driver release, and has to be at least one of the in-house tests they use to gauge performance of a driver. This led to the following idea...

    The Premise:
    nVidia has to dedicate time and support to testing 3DMark every release. Any substantial changes in performance should cascade outwards, especially negative performance.
    • If 3DMark drops measurably in performance. This would confirm larger, measurable, performance issues in titles with less focus could exist.
    • If 3DMark does not change in performance, other titles are not guaranteed to be reflective of this. More testing is needed then.

    So I currently only have used one benchmark, what do you think that means :)? This article may or may not be finished, I may go back further and testing more drivers.

    Testing Methodology:
    • Three runs per test.
    • Six runs per driver.
    • Scores below are averaged of three. GPU scores recorded and compared separately.

    Results (Charts):
    Blue Scores - Fire Strike Scores
    Green Scores - Fire Strike GPU Scores
    [​IMG]
    [​IMG]

    Results (Numbers):
    [​IMG]
    [​IMG]
    [​IMG]

    Results SLI (Charts):
    [​IMG]
    [​IMG]

    Results SLI (Numbers):
    [​IMG]
    [​IMG]

    [​IMG]
    [​IMG]

    Testing Notes:
    • 353.38 – A huge improvement over 353.30 in terms of stability and performance.
    • 353.30 - Just could not get a 3rd run on 3DMark Fire Strike Extreme tried 5 times even on stock w/ reboots. GPUs do not idle (go to a lower power state) once overclocked.
    • 353.06 - GPUs do not appear idle (go to a lower power state) once overclocked.
    • 352.86 - GPUs do not appear idle (go to a lower power state) once overclocked.
    • 350.12 - GPUs do not appear idle (go to a lower power state) once overclocked.
    • 347.88 - GPUs do not appear idle (go to a lower power state) once overclocked.
    • 347.52 - This is driver does not have the 20 seconds freeze in NVCP when changing 3D settings. GPUs idle (enter a lower powerstate) even with OC set!
    • 347.25 - 3DMark won't launch if Afterburner OSD is running. Changes to the NVCP 3D Settings don't seem to be sticking i.e. Prefer maximum performance, however scores are the highest.

    Overclocking Retested on the Latest Drivers (353.30):
    [​IMG]

    Overclocking Notes:

    Why is the 780 Ti struggling against a card it used to beat?
    A.) Well, we are EOL, I would assume that the 3 years of Kepler optimizations have already been made.
    • I expect new title support, but I don't expect the same level of focus on my GPU anymore.

    B.) Maxwell architecture is new.
    • Consistent newer title reviews have shown the GTX 980 / 970 are currently ahead of the GTX 780 Ti. Newer titles require newer drivers.
    • Most claims that the GTX 780 Ti used to beat the GTX 980 / 970 are made based on launch reviews. Older drivers were used.
    • New architectures mature with driver release.
    • Benchmarks and titles where Tessellation are heavily taxed, Maxwell has 3x tessellation performance as Kepler.
    • The GTX 970 and 980 also have another hardware advantage, another 1GB of VRAM, or 512 MB of VRAM in the 970 case.

    C.) GCN architecture isn't new!?
    • AMD has one architecture to support. Optimizations for the biggest GCN card trickle all the way down to the smallest GCN card, or even older cards.
    • Most claims that the GTX 780 Ti used to beat the R9 290X are based on old launch reviews. Older drivers were used. Newer titles require newer drivers.
    • All architectures mature with driver releases. Even GCN will be tapped out at some point in the future.

    D.) The transition to the 35x.xx family has left us a little down in terms of performance.
    • Windows 10 / DX12 driver changes are slowly being merged into 34x.xx family of drivers. Bugs will come up first, optimizations always come afterwards, if they come at all.
    • They could have done a better job in merging the branch changes in the drivers.

    E.) All The Console Ports Have Been Virtually Terrible
    • Batman: Arkham Knight – Sales have been stopped it was so bad.
    • Assassin’s Creed Unity – Appalling and terrifying release.
    • Watch_Dogs – Hatefully now known as Patch_Dogs.
    • VRAM requirements have tripled over night with no perceivable increase in image quality.
    • The architecture in the consoles GPUs is GCN. Games are being designed from the ground up to work well on AMD hardware.
    • Newer games simply will have a few more low level AMD optimizations at launch than nVidia ones.

    Conclusion (as of 7/3/2015):
    The transitioning into 35x.xx family has caused a certain amount of performance penalty to some Kepler users, for example a loss of 4.05% in GPU score in Fire Strike, and 3.88% Fire Strike Extreme. A loss though should be read in context. The loss in performance since at least driver 347.25. I don't think there is a huge conspiracy to "withhold" performance as people speculate, but there is enough data to merit a review by nVidia. I don't expect major performance gains any more, now that we are EOL, but we shouldn't be going down in performance in a benchmark that even pre-dates this video card.

    The best driver currently for Kepler, in terms of 3DMark performance I have tested thus far is 347.25. If you must use a 35x.xx driver, I highly recommend the 350.12 or the 352.86. I have updated the charts and added the actual numbers.

    Updated to Conclusion:
    The numbers since 340.52 only show a lost of 3.96% loss of performance in the GPU score, or 2.09% in overall score in Fire Strike. The latest hotfix driver 353.38 seems to be that nVidia is taking the issues serious and managed to resolve not only the performance issues but the vast majority of TDR crashing that was occurring.

    Also note: The 353.30 is the worst driver I have ever used from nVidia and I have had Quad SLI and GTX 590s before.

    Anecdotal Comments:
    I have greatly enjoyed my time with the GTX 780 Tis, even though they are not good overclockers. Out of all my nVidia experiences, they have been my favorite cards. nVidia driver stability has been usually good for me, that being said, the 353.30 is the worst driver I have ever used from nVidia and I have had Quad SLI and GTX 590s before.

    I have been using the 780 Tis since November 2013. I did my testing in my configuration because this has been my experience with Kepler GPUs in SLI. To me there hasn’t been a mass reduction in performance, I have never felt this crushing ominous crippling of performance that so many people are claiming is happening. I don’t doubt there are issues, issues which seem to vary based on each user’s configuration (including driver issues) but as far as I can tell no one has taken the time to prove many of these claims. I recommend everyone, if they can, to use the 347.25 driver, it is a solid performer and very stable. Maxwell also predates this driver which refutes the whole “nerfing Kepler to make Maxwell shine” theory. Looking at the trend up to the 347.25 driver, performance has been consistent. I personally do not have the TDR bug in Windows 8.1, I did have it in Windows 10, it does look like they are working on a fix finally though.

    The tests here are for a narrow premise and not all encompassing, I realize that. Again the point of this article is encourage neutral fact based discussion but also provide a reference to repeatable and reproducible tests. All my test results are under the user RagingCain at 3DMark Results, if you want/need that extra proof.

    My Driver Recommendations In Order
    • 347.25 – nVidia control panel settings do change, but when the page refreshes only the default choices appear to be selected.
    • 340.52 (For Non-SLI) – Performs slightly better in single GPU setups than 347.25, but not SLI.
    • 353.38 – Have not tested Shadowplay or Geforce Experience. No freezing loading the nVidia Control panel, settings do stick.
    • 350.12 – Had difficulties using Shadowplay.

    Article Source: ByteMeDev.com
     
    Last edited: Jul 3, 2015
  2. mahanddeem

    mahanddeem Master Guru

    Messages:
    663
    Likes Received:
    28
    GPU:
    Asus GTX1080Ti StrX
    Quality work mate. Very informative and thorough work.
    I myself don't believe in the "intentional" drop in Kepler performance and you proved it.
    Thanks.
     
  3. Barry J

    Barry J Ancient Guru

    Messages:
    2,803
    Likes Received:
    152
    GPU:
    RTX2080 TRIO Super
    respect good interesting and a lot of information
     
  4. zer0_c0ol

    zer0_c0ol Ancient Guru

    Messages:
    2,976
    Likes Received:
    0
    GPU:
    FuryX cf
    This is a shame, those cards were expensive.
     

  5. Genrix

    Genrix Member

    Messages:
    11
    Likes Received:
    0
    GPU:
    GTX 780 JS Palit@1228
    You're trying to find truth, but are looking wrong.
    The truth in ratio fps between Maxvell and Kepler during start sell Maxvell and fps now.
    Compare to Crysis3 / Battlefield 3 / Watch Dogs: Kepler 780 vs 980 vs 290x . What percentage difference?
    Now compare in Wither 3 and Project Cars. What percentage?
    Kepler did not slow down by NV, but NV not put optimizations for game in driver. Game is working as developer made it. With all faults in game code.
    Read this post _gamedev.net/topic/666419-what-are-your-opinions-on-dx12vulkanmantle/#entry5215019

    But AMD make it too. Compare 7970 vs 290x in games 2013, first half 2014: +30-35%.
    Then look games in second half 2014 and later - you can find some games where 290x have abnormal high performance: +50-70%.
     
    Last edited: Jun 28, 2015
  6. ---TK---

    ---TK--- Guest

    Messages:
    22,104
    Likes Received:
    3
    GPU:
    2x 980Ti Gaming 1430/7296
    Had the 2 exact cards for over a year and a half and haven't seen much of a difference in fps unless the new driver has specific optimizations for certain stuff. I mentioned more than once on this forum about hawaii being amd top card for 2 years and they squeezed every last drop of performance on these cards. In the same time frame 3 sets of maxwell cards were released. 750ti, 750 and 980 970 960 and titan x and 980ti. They had a lot of new stuff to support as well as older cards. Imo that is the reason 290x and 290 got better with time.
     
  7. moab600

    moab600 Ancient Guru

    Messages:
    6,658
    Likes Received:
    557
    GPU:
    PNY 4090 XLR8 24GB
    Kepler downgrade is real, 290 is pwning all keplers in the new batman game, where on previous nvidia heavily favored on batman games(770 defeated 290x).

    nvidia downgraded performance, does not matter how it's done it is done. i kinda felt ever since FC4, that mine 780 performing very bad even with my monster oc.

    Feel really sad for 680 770 owners, those cards are so poor, there was no reason to buy them over 280x and 7970.
     
  8. mahanddeem

    mahanddeem Master Guru

    Messages:
    663
    Likes Received:
    28
    GPU:
    Asus GTX1080Ti StrX
    I really can't understand how some think! why they feel the conspiracy theory everywhere? middle east syndrome:wanker:?

    So people who have 680, 770 or 780ti should expect their cards will perform ideally forever? or those who have 780ti should match and exceed 980 performance? well why? shouldn't newer games preform better on newer gpus and newer architecture?

    So because someone bought a 680 at release for 500$ they should still compete now in June 2015 with a 500$ card?
     
  9. evilos

    evilos Guest

    Messages:
    20
    Likes Received:
    1
    GPU:
    Gtx 970 G1
    compare 960 to 780 or 780ti in Witcher 3, game is not the same as at E3 2013, i cant see that amazing tesselation like everyone is saying, lots of textures are flat, water looks bad, plus is a console port and still 960 is at the same level as 780, when we move on to Project Cars, its even worse, 770/680 is destroyed by 960(1024/64/32/128bit) with gameworks, nvidia can control their gpus performance
     
  10. moab600

    moab600 Ancient Guru

    Messages:
    6,658
    Likes Received:
    557
    GPU:
    PNY 4090 XLR8 24GB
    You don't even own a 780 so u don't know what it all about, and no i don't expect 780ti to outperform 980, no a bit.

    I do expect 780ti to be better than 290x at all times or at least 90%, but seeing how 780ti was faster than 290x on release, 770 was faster than 280x and seeing today how same AMD cards that compete with 780 and 770, got better why Kepler got worse... doesnt make any sense.

    Just look at recent games, 280x is always better than 770, 290 always better than 780, 290x tops 780ti easily or ties it... and some paid 200$ more for 780ti, pff what joke.
     

  11. -Tj-

    -Tj- Ancient Guru

    Messages:
    18,095
    Likes Received:
    2,601
    GPU:
    3080TI iChill Black
    347.26, 353.06 (highest API calls until now) and 353.38 are the fastest drivers, I saw the same in raw brute force. And your chart shows it too..


    Btw you can't really come to conclusion based on firestrike, its too generic imo. Good job non the less. :thumbup:
     
  12. undeadpolice

    undeadpolice Master Guru

    Messages:
    203
    Likes Received:
    0
    GPU:
    EVGA GTX980 K|NGP|N (SLI)
    290x owners are also facing the same issue as AMD only improves on 390x with separate driver and leaving out 290x even through both have the same core.

    I wouldn't be surprise if AMD purposely gimp the drivers for 290x.
     
  13. SplashDown

    SplashDown Maha Guru

    Messages:
    1,131
    Likes Received:
    406
    GPU:
    EVGA 980ti Classy
    I know, I found a deal on 780ti, brand new for $300, and I couldn't pull the trigger. this is crazy really, I'm just having a hard time seeing the 970 beat a 780ti, because of a driver? On paper you'd think the 780 would smack it silly. I get same performance from 347.88 to 353.30, same numbers, no drop just same, so I guess that's all I get outta my 770.
     
    Last edited: Jun 28, 2015
  14. Undying

    Undying Ancient Guru

    Messages:
    25,207
    Likes Received:
    12,611
    GPU:
    XFX RX6800XT 16GB
    This. Indeed a good job but it would be nice to actually see some game benchmarks on Kepler cards on few different drivers. TK always saying how ridiculous this downgrade is, showing us some synthetic benchamrks but i never saw him posting/playing some games at all.

    All this year games are showing strange results on Kepler cards and i dont see this trend will change.
     
  15. moab600

    moab600 Ancient Guru

    Messages:
    6,658
    Likes Received:
    557
    GPU:
    PNY 4090 XLR8 24GB
    Kepler early EOL is depressing, seeing how performance went down and GCN counterpart rose up always bring me the question.

    Was Kepler doomed from beginning or nvidia playing dirty tricks? or both?
     

  16. Genrix

    Genrix Member

    Messages:
    11
    Likes Received:
    0
    GPU:
    GTX 780 JS Palit@1228
    Make optimized game ready driver for Maxwell and not doing this for Kepler - this is dirty trick. But every optimization cost money, because people has works.

    I see two way for fix it:
    1. Start forum threat for every new game with claims to Nvidia to optimize game driver. Maybe this help us. :)
    2. Create this optimization itself or with community.

    For 2: We can use GPU profiler, for ex Intel GPA. I has try to use Nsight, but I cant inject in final game. GPA make this inject very simple.
    And I dont know how I can replace any game code in runtime to good. Maybe we need create dx wrapper.

    3. Ask Nvidia give us tools to create optimizations. :)

    ps. Sorry for bad translate.
     
  17. ---TK---

    ---TK--- Guest

    Messages:
    22,104
    Likes Received:
    3
    GPU:
    2x 980Ti Gaming 1430/7296
    If you want Kepler gaming benches buy yourself a Kepler and bench some games. I am not here to pander to every request.
     
  18. Undying

    Undying Ancient Guru

    Messages:
    25,207
    Likes Received:
    12,611
    GPU:
    XFX RX6800XT 16GB
    Idk, i thought we are here to share benchmarks and opinions. You can prove some points and not only be talk.

    I guess me buying a Kepler card is easier then you actually do some gaming, meaning you already have two of them.
     
  19. ---TK---

    ---TK--- Guest

    Messages:
    22,104
    Likes Received:
    3
    GPU:
    2x 980Ti Gaming 1430/7296
    I share opinions all the time, to call me out because I did not provide game benches because you asked for them personally in another thread is a bit childish. Like I said I do not have to provide anything some random guy from the Internet asked for something. These cards are going to be replaced very soon anyway. I do not have the patience to do this kind of stuff like the OP did here. I shared my opinion of my same exact cards to the OP with the same exact cards.
     
  20. Singleton99

    Singleton99 Maha Guru

    Messages:
    1,071
    Likes Received:
    125
    GPU:
    Gigabyte 3080 12gb
    Nice post and a good read ,, thanks for your input into the Kepler crippling conspiracy.

    atm im at a loss as to where my next upgrade should be in terms of gpu , i want to go back to a single gpu and i do fancy the 980ti , but what will happen in a year from now with that card , will we have the same ,
    If it was ever 100% proved that they are and did cripple kepler ,,, well i don't think i would like to support such antics and my money would go elsewhere ...
     

Share This Page