What about Crysis on Very High?

So we tested Crysis on Very high settings with our single 9800 GX2 and Quad setup. Here’s what we got:




As we can see, with very high quality, performance starts to diverge between the dual and quad GPU NVIDIA solutions. It’s also interesting to note that performance doesn’t drop a great deal when moving up in quality. This indicates that the 9800 GX2 is still system bound in some way. And oddly enough, it looks like it is more system bound at the higher quality setting.

To test this absurd theory, we decided to see what happened when we overclocked our 8 CPUs from 3.2 to 4GHz. Let’s test the theory that we are CPU bound to about 45-50 fps at high quality and 40 fps at very high quality. Here is what our 25% overclock netted us when we tested with both very high and high quality using the 9800 GX2 in Quad SLI.


I’ll give you all a second to pick your jaws up off the floor....

Ok, break’s over. We see more than a 50% performance improvement per percent increase in CPU clock speed; a 25% clock speed increase netted us more than half that in real performance under Very High Quality settings. We saw less than 50% improvement at lower res for High Quality plus Very High Shaders until we hit 1920x1200, which netted us a 15% gain on a 25% increase in clock speed.

This indicates that the higher the graphical quality, the MORE CPU bound we are. Crazy isn’t it? It's counter-intuitive, but pure fact. In speaking with NVIDIA about this (they have helped us a lot in understanding some of our issues here), the belief is that more accurate and higher quality physics at higher graphical quality settings is what causes this overhead. Also, keep in mind that we are testing in a timedemo with AI disabled.

And that’s not where it ends. We are platform bound as well. Yes, I said platform bound. A quick check on 780i returned these results:


Not that we are still CPU bound here even though performance is about 25% higher than on Skulltrail (we benefit even more from a platform change than from an overclock). This is with the same speed CPU. It’s quite unfortunate that we stumbled across all of this last night with the help of NVIDIA while trying to troubleshoot our issues. Remember we said that NVIDIA expects a 60% improvement in Crysis at 19x12 with Very High Quality settings. The added number of cores, the fact that I’m only able to run two FB-DIMMS at the moment (for half of the system memory bandwidth I should have), the arrangement of the PCIe lanes on Skulltrail … All of this contributes to our system limitation here and our inability to see scaling from 9800 GX2.

Now, we have seen better performance on Skulltrail in the past, so it is unclear if there is something we can do to remove some of this system limitation at this point. We will certainly be exploring this further as we would still like to make a single platform work for all of our graphics testing. If we can’t then we’ll move on, but it is useful to discover if this is an Intel issue, an OS issue, a driver issue, or something else. If it’s fixable we need to find out how to fix it, as there are probably two or three people out there who’ve purchased a D5400XS board and will not be happy if it performs much worse than 780i boards in cases like this. My working theory right now is that I applied some hotfix or changed some seemingly benign OS setting that caused some problem somewhere. But like I said, we’ll track it down.

Overall Performance Scaling with 4 GPUs Do we see the same thing with World in Conflict?
POST A COMMENT

52 Comments

View All Comments

  • DDH III - Thursday, October 23, 2008 - link

    So you're saying that when my 1st 9800gx2 gets here in the mail next week, that it actually help my landlord on his heating bill? That's good because he pays for the electric too, and I feel kinda bad already. Since while heating bills in Interior Alaska are nasty, the electric is just as .. but anyways.

    Great review. Though I won't be able to run these in quad on my current mobo. But then, I never liked the FPS's much since quake, which brings me to my point:

    These days I play insane amounts of Supreme Commander. It is CPU intensive. If you are looking for another Benchmark when you start cranking up the GHz, ...well just give it a look.

    But to say a bit: 8 player is CPU pain. I only have dual cores, but my first dual cores were AMD (my games never needed the most rocking FPS) I had to buy into go C2D, just for game speed..forget the eye-candy.

    In short I read this article and went..huh...me too.

    Anyways, I am one of those Dell 3007 owner's, and I learned a lot. TY. And I don't live in North Pole AK, I only work there.




    Reply
  • luther01 - Friday, May 09, 2008 - link

    hi guys. i have a quad 9800 gx2 system as follows:

    q6600 processor @2.8GHz
    4 gig pc-6400 RAM
    780i motherboard
    2x 9800gx2

    i must say that so far i haven't noticed any performance gains over one 9800gx2, and in some games like The Witcher i have actually had a performance drop of 10-15 fps. In crysis, the frame rate in v high settings is disappointingly low, sometimes dropping to 15fps in large environments. i think maybe there is another bottleneck in my system
    Reply
  • Das Capitolin - Sunday, April 27, 2008 - link

    It appears that NVIDIA is aware of the problem, yet there's no word on what or when it might be fixed.

    http://benchmarkreviews.com/index.php?option=com_c...">http://benchmarkreviews.com/index.php?o...&tas...
    Reply
  • Mr Roboto - Sunday, March 30, 2008 - link

    Over at tweaktown they're coming up with the same results as Anand did even with the 9800GTX.
    Is it possible a lot of the problems are because of the G92\94 has a smaller 256-Bit memory interface? Could it cause that much of a difference especially combined with less VRAM. Just a thought.

    http://www.tweaktown.com/articles/1363/1/page_1_in...">http://www.tweaktown.com/articles/1363/1/page_1_in...
    Reply
  • ianken - Thursday, March 27, 2008 - link

    ...logging CPU loading shows it evenly loaded across four cores at about 50%. But then I'm running a lowly 8800GTS(640mb) SLI setup.

    The numbers in this review just seem kinda off.
    Reply
  • DerekWilson - Friday, March 28, 2008 - link

    what settings were you testing on?

    we see high quality and very high quality settings have an increasing impact on CPU ... the higher the visual quality, the higher the impact on CPU ... it's kind of counter intuitive, but it's a fact.

    you actually might not be able to hit the CPU limit with very high settings because the GPU limit might be below the CPU limit... Which seems to be somewhere between 40 and 50 fps (depending on the situation) with a 3.2GHz intel processor.
    Reply
  • seamusmc - Wednesday, March 26, 2008 - link

    I like Hard OCP's review because while the Quad SLI solution does put up higher numbers then a 8800 Ultra SLI solution they pointed out some serious problems in all games besides Crysis.

    It appears that memory is a bottleneck and many games have severe momentary drops in FPS at high resolutions and/or with AA, making the gaming experience worse then an 8800 Ultra SLI solution. I strongly recommend folks take a look at Hard OCP's review.

    AnandTech's review only covers average FPS which does not address nor reveal the kinds of issues the Quad SLI solution is having.
    Reply
  • B3an - Wednesday, March 26, 2008 - link

    Thanks for mention of the HardOCP review. A lot better than Anands.

    Very disappointed with Anand on this article. I posted a comment asking why i was getting such bad frame rates at high res with AA, and Anand did not even address this. When they must have run in to it at 2560x1600, and just about all the games they tested at 2560x1600 will get killed with AA because of the memory bottleneck. I'm talking from trying this myself. If i knew about it i wouldn't have got a GX2 as it's pretty pointless with a 30" monitor.
    So are they getting paid by NV or what?

    Very disappointed.
    Reply
  • AssBall - Wednesday, March 26, 2008 - link

    "Very disappointed with Anand on this article. I posted a comment asking why i was getting such bad frame rates at high res with AA, and Anand did not even address this."


    They are trying to answer their own questions and solve their own problems right now. This comment section is for comments on the article, not your personal technical support line to Anand.
    Reply
  • B3an - Thursday, March 27, 2008 - link

    When i said "I posted a comment asking why i was getting such bad frame rates at high res with AA, and Anand did not even address this."

    I meant, why has this website not addressed this issue in any of there GX2 articles. As any decent website would, especially being as it's not hard to run in to. I mean a high-end card like this would be for people with 24" and 30" monitors like myself. As anything lower than 1920x1200 a GTX would be more than enough. Yet the card has massive memory bandwidth problems and/or not enough usable VRAM. So frame rates get completely crippled when a certain amount of AA is used on games.
    Making the card pretty pointless.
    Reply

Log in

Don't have an account? Sign up now