We’ve received a number of reports over the last month of X-Plane hanging on startup with NVidia graphics cards on Windows.  Most users don’t see this, but some do.  We fixed one use of off-screen rendering in X-Plane 9.67 and this fixed some of the users with hung video cards but not all.

I am working with the remaining users now; my hope is that we’ll find something we can change in X-Plane to work around the problem.  I don’t believe that the code in X-Plane 9.62 was actually incorrect, but it was simple enough to change.

I have also read some posts regarding poor performance on the GeForce 400 series (Fermi cards) vs. the older 200 series.  I have no hard data on this, and frankly, a lot of the discussion on the net strikes me as completely speculative.

Fortunately alpilotx has a GeForce 500 series card on order, and he is working with me on next-generation DSF renders for X-Plane 10.  From my perspective, the big question about the GeForce 400-500 series (Fermi and beyond) is: how well do they implement the DirectX 11/OpenGL 4.0 feature set?  In particular, X-Plane 10 is going to make heavy use of full hardware instancing, and while this has been available on ATI cards since their HD2000 series, the implementation has been partly in software on NVidia DirectX 10 cards.

The question our users want to know is: if I have a preference for NVidia, can I get a DX11-class NV card for X-Plane, or should I move to ATI.  By looking at instancing performance, perhaps we can determine if these cards are contenders.

(I am not worried about overall “how many fps” do you get because y’all can measure that now with X-Plane 9.)

Edit: since a number of you have jumped in with performance reports: please post the following performance info in your replies:

  • Precise graphics card
  • CPU with clock speed
  • Results of –fps_test=3 (3 phases) with X-Plane 9.67.

About Ben Supnik

Ben is a software engineer who works on X-Plane; he spends most of his days drinking coffee and swearing at the computer -- sometimes at the same time.

26 comments on “X-Plane Hangs on Start with NVidia Cards

  1. Hi Ben,
    I have a EVGA Nvidia GTX 570 SuperClocked video card, I am not having any types of problems with X-Plane 9.67 running at 1920×1200.
    If you need any testing done let me know.

  2. My old GTX285 card produces more frames than my new GTX 580. I don’t know what the problem is but I have tried everything but still no gains in FPS. Last night I borrowed GTX 470 from a friend and did some testing but still the same poor performance. I’m running on 4ghz quadcore i7 .

  3. Hi Ben,
    I found how 🙂

    Here’s what I got:
    FRAMERATE TEST phase 1:time=30frames=1789fps=60
    HASP USB-key message: hasp_login: no key found
    FRAMERATE TEST phase 2:time=30frames=1799fps=60
    FRAMERATE TEST phase 3:time=30frames=1799fps=60
    Airport-Navigator v1.1.2 [INFO]: plugin disabled
    SimplePushback 0.5: Disabled
    DiscreetFPS: disabled
    Airport-Navigator v1.1.2 [INFO]: XPluginStop
    Airport-Navigator v1.1.2 [INFO]: AptNav::pluginEnd
    SimplePushback 0.5: Stopped

  4. ooops:
    FRAMERATE TEST phase 1:time=30frames=2604fps=87
    HASP USB-key message: hasp_login: no key found
    FRAMERATE TEST phase 2:time=30frames=3258fps=109
    FRAMERATE TEST phase 3:time=30frames=3589fps=120
    Airport-Navigator v1.1.2 [INFO]: plugin disabled
    SimplePushback 0.5: Disabled
    DiscreetFPS: disabled
    Airport-Navigator v1.1.2 [INFO]: XPluginStop
    Airport-Navigator v1.1.2 [INFO]: AptNav::pluginEnd
    SimplePushback 0.5: Stopped

  5. ATI 5870, 1GB
    MacPro 2.93GHz, 4cores
    2560x1600pix

    FRAMERATE TEST phase 1:time=30frames=2578fps=86
    FRAMERATE TEST phase 2:time=30frames=2735fps=91
    FRAMERATE TEST phase 3:time=30frames=2735fps=91

    regards
    arti

  6. Radeon HD 3870 512 MB GDDR4 @ 1680×1050
    MacPro3,1 – 2.8 GHz (5400-series Xeon) quad-core x2
    FRAMERATE TEST phase 1:time=30frames=1643fps=55
    FRAMERATE TEST phase 2:time=30frames=1756fps=59
    FRAMERATE TEST phase 3:time=30frames=1759fps=59

    The GPU performance seems decent but I’d like more objects. I run it with ‘a lot’ of object with medium detail distance. It really struggles in a city when it’s cloudy with a modern aircraft model like the x737 v4.

    I suppose a new graphics card won’t increase the object and cloud performance much if at all? The new i7-2600 quad only benchmarks equally as fast as my 8-core machine at multithreading but is almost twice as fast with a single thread. It doesn’t seem like much of an upgrade, all things considered.

  7. Nvidia GeForce GTX 480
    Dual Xeon X5680, 4GHz, 12 cores, 24 threads

    FRAMERATE TEST phase 1:time=30frames=4196fps=140
    FRAMERATE TEST phase 2:time=30frames=4622fps=154
    FRAMERATE TEST phase 3:time=30frames=5096fps=170

  8. nVidia GTX 485
    Core 2 Extreme QX6700 2.66GHz (4 cores)
    8GB RAM
    Windows 7 64-bit

    FRAMERATE TEST phase 1:time=30frames=1484fps=49
    FRAMERATE TEST phase 2:time=30frames=1498fps=50
    FRAMERATE TEST phase 3:time=30frames=1584fps=53

    – Robin

    – RObin

    1. OOOPS – delete my stats here. I was dreaming that I had a GTX485 … really it’s a GTX285! Must get a new card next week … or a whole new PC . Or a Mac Pro …

      – Robin

  9. Hey, Ben,

    I tried to run the FPS test again with max settings (which run fine at Innsbruck) but it won’t perform the test anymore. I successfully did it once as you can see above, so I wanted to see the results with max settings. Now, it seems the command line options don’t work. I’ve tried both single and double dashes, and different tests, but the commands simply have no effect. Any ideas?

    Regards, Patrick

  10. Update: It seems the Cmd options are working, so something else must be wrong. Excerpts from log.txt:

    X-System folder:’E:\X-Plane 9/X-Plane 9/’, case sensitive=0

    ‘-fps_test’=’3’
    ‘-require_fps’=’20’
    OpenGL Vendor : NVIDIA Corporation
    OpenGL Render : GeForce GTX 460/PCI/SSE2
    OpenGL Version : 4.1.0 (410)

    Preload time: 1780328.
    Total load time: 21
    HASP USB-key message: hasp_login: no key found
    HASP USB-key message: hasp_login: no key found
    HASP USB-key message: hasp_login: no key found

    The first one continued past that. I’m not sure what’s going on.

    — Patrick

  11. Hi Ben,

    AMD Phenom(tm) 9850 Quad-Core Processor @ 2511.636
    4GB RAM
    Gentoo Linux
    GTX470 Zotac AMP (http://www.zotac.com/index.php?option=com_wrapper&view=wrapper&Itemid=100062&lang=ca)
    nvidia driver: 270.18

    FRAMERATE TEST phase 1:time=30frames=2044fps=68
    FRAMERATE TEST phase 2:time=30frames=2145fps=71
    FRAMERATE TEST phase 3:time=30frames=2269fps=76

    I have moved from 8800GTX and did not notice too much of a difference. ;|

    I am very glad to hear that there is a room for improvement.

    Kind regards,
    Kamil

  12. Phenom II X4 @ 3.8GHz, Radeon HD4770
    Ubuntu 10.04 64bit, 2/11 Catalyst drivers, XP9.67

    1024×768
    FRAMERATE TEST phase 1:time=30frames=3206fps=107
    FRAMERATE TEST phase 2:time=30frames=3647fps=122
    FRAMERATE TEST phase 3:time=30frames=4171fps=139

    1920×1200
    FRAMERATE TEST phase 1:time=30frames=3136fps=105
    FRAMERATE TEST phase 2:time=30frames=3556fps=119
    FRAMERATE TEST phase 3:time=30frames=4060fps=135

  13. Plugin problem? If you remove third party plugins does that stuff go away? (If it does you’ll need to contact the plugin author.

  14. model name : Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
    cpu MHz : 2801.000
    MemTotal: 4110292 kB
    S.O = Ubuntu 10.10
    ‘–fps_test’=’3’
    OpenGL Vendor : NVIDIA Corporation
    OpenGL Render : GeForce GTX 275/PCI/SSE2
    OpenGL Version : 3.3.0 NVIDIA 270.29 (330)
    FRAMERATE TEST phase 1:time=30frames=4372fps=146
    FRAMERATE TEST phase 2:time=30frames=4336fps=145
    FRAMERATE TEST phase 3:time=30frames=4417fps=147

  15. model name : Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
    cpu MHz : 2801.000
    MemTotal: 4110292 kB
    S.O = Ubuntu 10.10 32 bit and 64bit
    ‘–fps_test’=’3’
    OpenGL Vendor : NVIDIA Corporation
    OpenGL Render : GeForce GTX 275/PCI/SSE2
    OpenGL Version : 3.3.0 NVIDIA 270.29 (330)
    FRAMERATE TEST phase 1:time=30frames=4372fps=146
    FRAMERATE TEST phase 2:time=30frames=4336fps=145
    FRAMERATE TEST phase 3:time=30frames=4417fps=147
    1920×1080

  16. HELLO GUYS, I HAVE TO BUY A NVIDIA GTX 460 with 768 MB of RAM MEDIAWORLD DDR5.
    WITH FSX performance is excellent, OR DO NOT GO ON MEDIA AND MICROSOCATTI 60 FPS with peaks of 100, the SET WITH MEDIUM-HIGH.
    HOWEVER, when I started X-Plane 9 HO WAS NOTED THAT THE SIMULATOR TO THE MAX 20 FPS, I HAVE TRIED TO WRITE THE KEY PREFERENCES, but nothing changes.
    Then I did a test I removed the NVIDIA MIA, I HAVE REPLACED MY OLD-Vapor-X ATI 4870 DDR5 2GB of RAM.
    I DID THAT I WAS RESULTING FSX microswitch, WITH A MAX FPS arriving at 40.
    When booting X-PLANE WAS Smoothly 9.
    Now there I ask this question, is it true that X-PLANE is optimized for FSX FOR NVIDIA AND ATI?

    HOW peformance is more powerful than the GTX 460 CHE THE VAPOR-X 4870.

  17. I already posted above but wanted to show results with SLI. Just FYI.

    OpenGL Render : GeForce GTX 480/PCI/SSE2
    OpenGL Version : 4.1.0 (410) (ForceWare 270.51)
    Dual Xeon X5680, 3.46GHz, 12 cores, 24 threads

    Single GPU
    FRAMERATE TEST phase 1:time=30frames=3861fps=129
    FRAMERATE TEST phase 2:time=30frames=4227fps=141
    FRAMERATE TEST phase 3:time=30frames=4572fps=152

    Force alternate frame rendering 1
    FRAMERATE TEST phase 1:time=30frames=2173fps=72
    FRAMERATE TEST phase 2:time=30frames=2247fps=75
    FRAMERATE TEST phase 3:time=30frames=2397fps=80

    Force alternate frame rendering 2
    FRAMERATE TEST phase 1:time=30frames=2324fps=77
    FRAMERATE TEST phase 2:time=30frames=2409fps=80
    FRAMERATE TEST phase 3:time=30frames=2527fps=84

    1. Interesting…it makes sense that AFR rendering would have poor performance – X-Plane has (as far as the driver is concerned) inter-frame dependencies. (This is more because OpenGL doesn’t provide a simple clean way to break these dependencies that I am aware of than because of X-Plane’s design) so there’s a lot of data transfer going on between the two cards.

      Also, we’re not usually fill-rate bound in X-Plane 9 on high quality DX11 cards, so the second card is probably not necessary.

      1. I just wanted to post it since I saw that no one else has. I didn’t expect better frame rates in X-Plane since forcing AFR for other games that didn’t have an SLI profile also resulted in lower frame rates. During the test, GPU usage is at a constant 24% and CPU usage is at 7%(almost 2 threads) so I wouldn’t expect SLI to be beneficial at this point anyway. I can however set Anti-Aliasing to 64x(overkill, I know) and Transparency Anti-Aliasing to 16x super-sampled with both GPUs at 43% usage and still get over 80FPS. I have found that turning on FSAA in X-Plane yields greatly lowered framerates compared to turning FSAA off in X-Plane and overriding anti-aliasing in the Nvidia Control Panel.

        I’m really looking forward to all the improvements in X-Plane 10. Thank you for all your hard work!

  18. Hi.

    I’m using a Nvidia GeForce 9800GT and when using a Matrox Triplehead2go, X-Plane hangs on start-up.

    I’ve updated and downloaded every new driver possible (the triple screen looks amazing). But X-plane hangs.

    When I revert back to single screen (original setup), X-plane loads.

    Anyone got any ideas what I could try?

    Thanks.

    Hunnas

Comments are closed.