When Things Go Wrong & The Test

It's worth noting that at one point this article had a very different tone to it, based on the benchmark results we had gotten. We recently replaced the test rig we run the PhysX articles on with a newer machine, and as the component of note, threw in an ATI Radeon X1900XTX as it's generally the fastest single-slot card we have that isn't an SLI card (i.e. the GeForce 7950 GX2). City of Heroes/Villains is a game we long ago established was CPU limited, so the choice in video cards is largely academic, or so we thought.

City of Villains Performance


It turns out that ATI's latest drivers have a problem with City of Heroes/Villains where the performance of the game chokes when using some of the advanced rendering features. Fortunately, we caught this issue, but for a while we were wondering why the PhysX card wasn't helping as much as expected. It's always interesting to discover where the bottlenecks are in different benchmarks. City of Heroes is a great testing components since it's an OpenGL title that isn't built on the Doom3 engine. Unfortunately, this is bad timing for ATI, given their recent OpenGL improvements on games that do use the Doom3 engine.

Due to the issues with the ATI card, we switched to testing with a 7950 GX2 instead. Here are the details of our test setup.

PhysX Testbed Configuration
CPU: Intel Core 2 Extreme X6800 (2.93GHz/4MB)
Motherboard: Intel D975XBX (LGA-775)
Chipset: Intel 975X
Chipset Drivers: Intel 7.2.2.1007 (Intel)
Hard Disk: Seagate 7200.7 160GB SATA
Memory: Corsair XMS2 DDR2-800 4-4-4-12 (1GB x 2)
Video Card: NVIDIA GeForce 7950GX2
Video Drivers: NVIDIA ForceWare 91.33
OS: Windows XP Professional SP2


Index PhysX Performance
Comments Locked

31 Comments

View All Comments

  • Calin - Friday, September 8, 2006 - link

    MMX and its successors were used to accelerate the hell out of many photoshop plug-ins (P4 in its worst days was faster running the optimised routines than Athlon64, and in many cases it was faster by a big amount).
    I think the video cards could be better at offloading this kind of calculation - maybe even more optimised routines will come soon (in many cases, graphic professionals use top-of-the-line cards, or even workstation-builds like NVidia Quadro and ATI FireGL)
  • Calin - Thursday, September 7, 2006 - link

    Using Ageia PPU gives you about 35% or more extra performance in CoV, or 6fps. Using a dedicated sound card instead of a lowly integrated sound will give you 6fps in benchmarks of the Quake3 or so engines, for a 5% or so difference in frame rate. For just this purpose, PPU is better than a dedicated sound card (even if more expensive)
  • DerekWilson - Thursday, September 7, 2006 - link

    I can find a more expensive audio solution for you :-)

    But this is true -- percent difference is what we need to look at here.
  • Calin - Friday, September 8, 2006 - link

    You could find a more expensive audio solution certainly - but I don't think you would be able to reduce the frame rate as compared to a $200 Creative 7.1 Channel Sound Blaster X-Fi Fatal1ty FPS (the price from Anandtech's own RealTime Pricing).
    Anyway, thanks for the article - nice written, and interesting. Thumbs up!
  • Marlin1975 - Thursday, September 7, 2006 - link

    How abotu trying the PhysX on a board that has intergrated graphics? Would it help those stuck with pos intel onboard or even the better Ati/Nvidia onboard graphics? Onboard usually covers 2d ok and has some 3d, but maybe with a little help onboard can move up with little cost. (cost as in when the PhysX comes down to the real world in pricing)?
  • Lonyo - Thursday, September 7, 2006 - link

    Also, for less than the price of a PPU you can get a 7900GT.
    If you want to try and suggest a PPU is better than a 7900GT as an upgrade, well, hmm.
  • PrinceGaz - Thursday, September 7, 2006 - link

    If a game is totally GPU limited (as is usually the case with onboard graphics), then a physics-card will make no difference whatsoever to performance.

    The PhysX card offloads work from the CPU, it does not offload any work from the GPU, so the PhysX card will not help people using onboard/slow graphics solutions.
  • Calin - Thursday, September 7, 2006 - link

    Also, a physics card "creates" more debris, which is not only physics intensive to compute (paths and so on), but GPU intensive to render.
    Anyway, integrated graphics usually reduce the quality and resolution of possible gaming - using the money for the physics card for a new (or additional) graphic card would be the cheapest solution to fast, quality gaming. Not to mention you could get multimonitor capabilities in the price, maybe DVI and so on.
  • DerekWilson - Thursday, September 7, 2006 - link

    This is quite true ... if you don't have a gpu that can handle what you are already throwing at it, a PhysX card won't do much. Sure it'll take off some CPU load, but chances are you aren't cpu limited. And if you tried turning up the debris settings you'd just be adding to the load on the GPU. Which could cause some performance decrease.

    We will keep this in mind for future tests and try to address the issue later. For now, it's safe to assume that you'll need at least a midrange quality graphics board to gain anything from PhysX.
  • yyrkoon - Thursday, September 7, 2006 - link

    I think you'd have to be insane to think that running a game on a system with onboard video is going to do well in a Physics intensive game . . .

Log in

Don't have an account? Sign up now