A little over a year ago, we first heard about a company called AGEIA whose goal was to bring high quality physics processing power to the desktop. Today they have succeeded in their mission. For a short while, systems with the PhysX PPU (physics processing unit) have been shipping from Dell, Alienware, and Falcon Northwest. Soon, PhysX add-in cards will be available in retail channels. Today, the very first PhysX accelerated game has been released: Tom Clancy's Ghost Recon Advanced Warfighter, and to top off the excitement, ASUS has given us an exclusive look at their hardware.

We have put together a couple benchmarks designed to illustrate the impact of AGEIA's PhysX technology on game performance, and we will certainly comment heavily on our experience while playing the game. The potential benefits have been discussed quite a bit over the past year, but now we finally get a taste of what the first PhysX accelerated games can do.

With NVIDIA and ATI starting to dip their toes into physics acceleration as well (with Havok FX and in-house demos of other technology), knowing the playing field is very important for all parties involved. Many developers and hardware manufacturers will definitely give this technology some time before jumping on the bandwagon, as should be expected. Will our exploration show enough added benefit for PhysX to be worth the investment?

Before we hit the numbers, we want to take another look at the technology behind the hardware.

AGEIA PhysX Technology and GPU Hardware


View All Comments

  • DerekWilson - Friday, May 5, 2006 - link

    We will be taking a look at CellFactor as soon as we can Reply
  • Egglick - Friday, May 5, 2006 - link


    It seems most likely that the slowdown is the cost of instancing all these objects on the PhysX card and then moving them back and forth over the PCI bus and eventually to the GPU. It would certainly be interesting to see if a faster connection for the PhysX card - like PCIe X1 - could smooth things out....

    You've certainly got a point there. Seeing as how a physics card is more like a co-processor than anything else, the PCI bus is probably even more of a limitation than it would be with a graphics card, where most of the textures can simply be loaded into the framebuffer beforehand.

    I still believe that the best option is to piggyback PPU's onto graphics cards. Not only does this allow them to share the MUCH higher bandwidth PCIe x16 slot, but it would also mean nearly instant communication between the physics chip and the GPU. The two chips could share the same framebuffer (RAM), as well as a cooling solution. This would lower costs significantly and increase performance.
  • DerekWilson - Friday, May 5, 2006 - link

    combo boards, while not impossible to make, are going to be much more complex. There could also be power issues as PhysX and today's GPUs require external power. It'd be cool to see, and it might speed up adoption, but I think its unlikely to happen given the roi to board makers.

    The framebuffer couldn't really be shared between the two parts either.
  • Rolphus - Friday, May 5, 2006 - link

    On page 2: "A graphics card, even with a 512-bit internal bus running at core speed, has less than 350 Mb/sec internal bandwidth." - er, I'm guessing that should read 350Gb/sec? Reply
  • JarredWalton - Friday, May 5, 2006 - link

    Yes. Correcting.... Reply
  • Rolphus - Friday, May 5, 2006 - link

    Thanks for the quick response - I've just finished the article. It's good stuff, interesting analysis, and commentary and general subtext of "nice but not essential" is extremely useful.

    One random thing - is struggling, or is my browser just being a pain? I've been having trouble seeing a lot of the images in the article, needing various reloads to get them to show etc.
  • ATWindsor - Friday, May 5, 2006 - link

    Anandtech images doesn't work properly if you disable referer logging (pretty annoying), can that be the root of your problem? (adblock disabling it or something) Reply
  • JarredWalton - Friday, May 5, 2006 - link

    Seems to be doing fine from our end. If you're at a company with a firewall or proxy, that could do some screwy stuff. We've also had reports from users that have firewall/browser settings configured to only show images from the source website - meaning since the images aren't from, they get blocked.

    As far as I know, both the images and the content are on the same physical server, but there are two different names. I could be wrong, as I don't have anything to do with the system administration. :)
  • Rolphus - Friday, May 5, 2006 - link

    Weird, seems to be fine now I've disabled AdBlock in Firefox... that'll teach me. It's not like I block any of AnandTech's ads anyway, apart from the intellitxt stuff - that drives me NUTS. Reply
  • JarredWalton - Friday, May 5, 2006 - link

    Click the "About" link, then "IntelliTxt". You might be pleasantly surprised to know it can be turned off. Reply

Log in

Don't have an account? Sign up now