Slashdot Mirror
Official Doom 3 Benchmarks Released
Rogerpq3 writes "Before the game goes on sale, id Software has been kind enough to release some benchmarks for DOOM 3 with the latest video cards on the market from NVIDIA & ATI. HardOCP has published the five page article which should help anyone trying to decide if they should upgrade their video card for DOOM 3. There's also an introductory note from John Carmack, mentioning: 'The benchmarking was conducted on-site, and the hardware vendors did not have access to the demo before hand, so we are confident that there is no egregious cheating going on.', and the HardOCP writers comment: 'As of this afternoon we were playing DOOM 3 on a 1.5GHz Pentium 4 box with a GeForce 4 MX440 video card and having a surprisingly good gaming experience.'"
Of course Nvidia's card is going to do better. Doom3 has a specialized codepath for nvidia hardware, while the ATI card does not.
If a codepath were written for the X800 series of cards, I'm sure the scores would be closer to each other.
I take the superiority of one card over the other with a grain of salt.
/^[A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,4}$/i
How about some benchmarks for a card I actually have, like a ti4800? ;-) Saying "suprisingly good gaming experience" on a GF4MX means nothing... are you seeing a creepy title screen and playing a pong minigame, or actually seeing 30fps+?
Sorry, but dropping $500 on a video card is just not an option, this would be more useful if we had some everyday specs.
Don't think of it as a flame---it's more like an argument that does 3d6 fire damage
they have only 3 cards listed in the test and none of htem are widely in use
Keep R'ing TFA -- they test (1) nVidia Geforce6800 Ultra (1st place with a bullet), (2) nVidia Geforce6800GT (strong second), (3) ATI X800XT-PE (3rd and more costly than (2)), (4) and (5) nVidia GeForceFX5950 and ATI9800XT (pretty much a tie -- ATI is a tad faster with AF [anisotopic filtering] but no AA [anti-aliasing], add in the AA and nVidia edges ahead.)
That's five, and at least two of them are what I'd call "widely in use." YMMV.
everything in moderation
Last I heard/saw, the 6800 still needed 2 molex connectors, and took up two expansion slots, sounded like a jet engine, and required a minimum 400 watt power supply. The ATI card uses much quieter cooling, requires one slot, and one power connector. For a machine that's on 24 hours a day in the same room I sleep in, noise is a big factor. If I needed the caliber performance of the latest/greatest card, and had an extra 400 to spend on a video card who's price will most likely be half that in 6-8 months, the extra $100 would be worth it, when factors other than an extra 3-5 fps come into play.
That's not to say I don't respect Nvidia, I swapped out a Radeon 9700 pro for a GF4 Ti4200 in this box, because the linux drivers from Nvidia gave me slightly better performance and much more stability in most of the OpenGL apps I'm running.
Which brings up another question- how will the native linux version of Doom III compare to it's windows counterpart?
If you're running the most recent CPU/GPU with a $hitload of RAM.. you're going to have a good gaming experience
WELL NO SHIT! What did you expect? The game to only run acceptably on hardware that doesn't exist yet? Geez..
As of this afternoon we were playing DOOM 3 on a 1.5GHz Pentium 4 box with a GeForce 4 MX440 video card and having a surprisingly good gaming experience
Why no benchmarks of this? IMO much more useful than a benchmark of a P4 3.6GHz system with 4GB of RAM and a 6800 Ultra..
I am the maverick of Slashdot
Just as an aside I picked up a geforce 6800 non-ultra last month and could not be more pleased with its performance - and it requires only one slot and a "standard" 300W power supply.
Are you implying that Carmack made the above statement? Because...he didn't. That's Kyle Bennet, the author of the HardOCP article speaking. Carmack only made the brief statement at the beginning (it's color coded to help you spot it), which states that "all of the modern cards play the game very well," and "there is no egregious cheating going on," and most importantly, "Nvidia drivers have been tuned for Doom's primary light/surface interaction fragment program."
I don't think Doom3 will be significantly changed to help out ATI, but I'm positive ATI will change their drivers to help out Doom3's performance. As Carmack pointed out, the Nvidia drivers have already been fine tuned for Doom. My guess is that ATI, after the fiasco with releasing the Doom alpha, hasn't had as much opportunity to optimize for Doom.
On the other hand, it's no surprise to see ATI losing to a card that obviously has more horsepower. Frankly, I'm impressed that a card that's so much cooler, smaller, and quieter does so well against Nvidia's monster. But in this case, at least, we see Nvidia's power fully utilized. Hopefully, ATI gets so more performance out of theirs, though.
-Dan
Just a couple of quick notes. First off, you must bear in mind that all of the cards they tested are DX9 capable, this is going to help out framerates quite a bit. In fact, when you look at benchmarks for the newer games, framerates drop off quite a bit when you start looking at cards like the GeForce 4.
:-D
More importantly, the boxes they did the benchmarking on were maxed out with specs like 2GB of DDR400 and an Athlon 64 or comparable processor. Unless you've got all the other specs to match the test box, you're looking at the best possible framerates you can get under the very best possible conditions on those systems. In addition to that, they had anti-aliasing turned off for several of those benchmarks.
Now compare those 60-70fps on that kind of box with whatever setup you've got...then swap out the video card for a GeForce 4Ti 4X00 and you're looking at maybe25-30fps with medium effects at 1024x768. That's almost unplayable.
Granted, I'm doing a good bit of guessing here, but this comes from a number of years of experience playing the latest games on older hardware. The basic sys-req's for the game are a GF3 or better - we can interpret that to mean it'll give you about 25fps at 800x600 with all the eye candy turned off if you're sporting a top-of-the-line GeForce 3. I doubt you're going to see good performance out of Doom 3 without anything better than a GeForce FX 5600.
Luckily, we'll all find out in a little less than two weeks
"It is seldom that liberty of any kind is lost all at once." -David Hume
Did you set everything to maximum or did you lower it? I found that terrain could be on high or normal and everything else needed to be normal except characters. Characters needed to be on low.
-]Phreak Out[-
Awwww, the wide-eyed innocents are posting! Look at this one-- he actually believes that one of the graphics card manufacturers is NOT cheating! It's tough life-lesson time, kiddo-- they BOTH have a history of cheating. Also, there is no Santa Claus or Tooth Fairy, and the Easter Bunny was paid $5M by ATI to optimize his egg-rendering scheme for their hardware.
Nvidia Cheating
ATI Cheating