id Software's Official DOOM3 Benchmarks

Any idea how this will effect our overclocks? Just curious, because my $299 BFG GT OC card I got from Bestbuy has been running fine at 415/1.05. If it fails to work at that setting for D3, will this mean I need to solder in that second power connector to bring it closer in line with a Ultra? Is it a matter of cooling, or a matter of extra power needed? I say this because NVidia made their comments about us not to expect the same OC' ability with D3 as with other games. I am wondering now if I should back off the OC settings to the factory default, so I do not burn out my card when running D3 for the first time if it is a matter of heat as opposed to power.
 
Ghosthunter said:
True but if your FPS is too low you wont have that complete gaming experience. Why would I not want to have an extra 20 FPS?
Click to expand...

i wonder is doom 3 still limited to 60 fps answer that cause this isnt quake 3 where fps actually mattered this is a completely different game and engine will you notice a difference between 40 to 60 prolly not. will you still have the same experence as on the faster hardware with the slower yes cause the gfx card is still going to do its job desplaying gfx on your monitor
 
Ghosthunter said:
Does Half Life 2 even matter? I doubt that game can even compare to Doom 3 graphicwise anyway.
Click to expand...


:rolleyes:

Nah halflife 2 doesn't matter /Fanboy mode: (if ati runs it faster that is) :rolleyes:
 
Ghosthunter said:
Does Half Life 2 even matter? I doubt that game can even compare to Doom 3 graphicwise anyway.
Click to expand...
The ancient HL2 benchmarks with a NV30 vs R300 that he is implying are outdated anyways. Even if the x800 is faster (which it probably will be), it won't be that much faster.

It wouldn't matter if a NV40 and R420 were compared today since the game is nowhere near ready to ship. By the time HL2 is released, the next generation of cards will be near introduction.
 
Aztlan said:
i think you are wrong on that

digit-life has a review of both cards running the hl2 beta

http://www.digit-life.com/articles2/gffx/nv40-3-p4.html

and the cards are neck on neck

anyway, on topic, like most of you I was quite amazed at the performance difference, specially with the GT beating the XT-PE
wow
Click to expand...

It's only a beta test. They are optimizing in favour of X800 cards. The noise and power consumption is the worst turn off with Geforce cards.
 
You know the ATI guys/gals are working their asses off trying to get better OpenGL performance right now. Having recently bought my first ATI card (just a 9600xt), I hope they pull it off, and soon...
The debate doesn't effect me too much though, I don't have the money for any of those high-end cards...
 
Nv just pwned ATI :eek:

Doom 3 engines will probably power games for the next 3-4 years, just like Quake 3 engines have been in tons of games.

I think I just saw the 6800 GT's just now sell out everywhere!
 
*DISCLAIMER - I only read first few posts, this post may seem idiotic, uninformed, or otherwise be flamebait.*

In my opinion, they showed what we've been saying with those tests, that a minimum system will barely get by in the game.
 
Chris_B said:
Anyone remember this http://www.hardocp.com/article.html?art=NDc0

Interesting to note the results are nowhere near these despite carmacks estimates.
Click to expand...
There are probably big differences between the "test2" and "demo1" benchmarks. Plus there's a 12 month difference in Doom3 development time. I wouldn't exactly call those benchmarks "carmack's estimates." That was the performance of the tested GPUs at the time the benchmarks were run.

Even considering that, the 5900U score was inflated of course since it was run in reduced precision, using the NV30 path that was dropped several months ago.
 
Xander2None said:
It's only a beta test. They are optimizing in favour of X800 cards. The noise and power consumption is the worst turn off with Geforce cards.
Click to expand...

The 6800 GT is a single slot, single molex solution. It doesn't use any more power than an XT PE. I'd bet that it isn't any louder, either.

Poor DOOM3 and OGL performance is the worst turn off with X800 cards.
 
If you ask me Im thinking of just hold out for the next gen cards lol. Neither generation of ati or nvidia can play this game at 1600x1200 with good aa/af settings on max quality. Just wait for the 512mb versions if you really want to play doom3 in all its glory.
 
I really dont get why some people are acting like this was totally unexpected, with ultrashadow and nvidia's reputation of being better in open gl it was safe to assume the nv cards would pull ahead by some degree. Nvidia are usually ahead in opengl whereas ati are usually ahead in d3d, big whoop dee doo. :rolleyes:
 
Shane said:
If you ask me Im thinking of just hold out for the next gen cards lol. Neither generation of ati or nvidia can play this game at 1600x1200 with good aa/af settings on max quality. Just wait for the 512mb versions if you really want to play doom3 in all its glory.
Click to expand...

The 6800 series gets ~60fps with everything maxed (except AA) at 1600x1200.

20040722092313_21506.gif
 
pxc said:
:rolleyes: if you have to tell yourself that...
Click to expand...


Something wrong with you? It was totally unoptomised, not even a beta just something valve threw together to showcase some of the stuff in the game. If we're gonna go by leaked shit then i can whoop and holler about getting 70fps in the doom 3 leak at 1280x1024, yeah its real representative of a finished product. :rolleyes:

Or if you wanna go on about it check the doom 3 results from what carmack said he "believes it to be representative of the final game in a performance sense" right here http://www.hardocp.com/article.html?art=NDc0. Funny i dont see the 9800 or fx boards getting those frame rates now, whys that i wonder? Could it be the game wasn't finished? SURELY NOT??? :eek: :rolleyes:
 
SHEESH, YOU'VE made this your religion haven't u? Kinda like the Muslims out to kill the infidels:
nVidiot: Human being brainwashed by false statements and propaganda from nVidia Corp. into believing all things GeForce are of the highest of superiority and defy all logical grounds of video cardness. Belief in 3Dmark2001 Benchmark scores regardless of quality, functionability, and compatability is a common worship and religion of these fundamentalists. However do not take these poor souls lightly they are armed heavily with propeganda, false statements, DetonatorXPs, and GeForces. Remember nVidiots come in all shapes and sizes. Be on guard or you will be assimilated.
Click to expand...
 
BoogerBomb said:
Does nobody here play with 9800Pro's and 512 RAM? DOes Id software honestly think that a common configuration is 2 gigs of ram and a $500 video card? I work in retail and bring home scratch. A 9800XT is STILL $300 which is now considered low end budget card from the way the article sounds. With my setup I might as well not even bother paying $60 for this game when it comes out if I have to turn off all eye candy to be able to play it.

Would have been nice to see benchmarks made on a real machine and not the dream machines that were used.
Click to expand...

Hear.... hear! Amen brother....

AMDXPM-2400+@2.3GHz/512MB/9800Pro->9800XTmod....

I want some eye candy!
 
oozish said:
SHEESH, YOU'VE made this your religion haven't u? Kinda like the Muslims out to kill the infidels:
Click to expand...


Amazingly enough ive had this sig since i signed up here 3 years or so ago because at the time it was very true, i dont post here as much now so i havn't bothered to change it. Now back on topic if thats ok with you of course. :rolleyes:
 
People are talking about how 'no one has gotten the code so no one could have optimzied for it' But it seems people are forgetting something.

Everyone except ID, and they did.. Carmak talked about having to put in a special code path for NV cards to make them work with Doom3 because their perf without it was crap. I want to see Benchmarks without NVOptimized code, having the engine optimized invalidates the game as a Benchmark of the cards.

If without the optimized code NV still beats out ATI then sure NV is faster in OpenGL (which is used by how many engines these days?) Anyone who didn't know that NV was faster in OpenGL must have lived under a rock, BUT when the engine has optimized paths for one card it becomes slanted. So is NV really faster in OpenGL anymore? or is it purely optimized engine code that makes NV faster... Hopefully we'll know sooner rather than later.
 
Ghosthunter said:
Does Half Life 2 even matter? I doubt that game can even compare to Doom 3 graphicwise anyway.
Click to expand...
Does HL matter? You know your family? Yes them. Its more importain then them! hehe :D
 
colin said:
You know the ATI guys/gals are working their asses off trying to get better OpenGL performance right now. Having recently bought my first ATI card (just a 9600xt), I hope they pull it off, and soon...
The debate doesn't effect me too much though, I don't have the money for any of those high-end cards...
Click to expand...

preech on brother neiter do i :)
 
Chris_B said:
Something wrong with you? It was totally unoptomised, not even a beta just something valve threw together to showcase some of the stuff in the game. If we're gonna go by leaked shit then i can whoop and holler about getting 70fps in the doom 3 leak at 1280x1024, yeah its real representative of a finished product. :rolleyes:

Or if you wanna go on about it check the doom 3 results from what carmack said he "believes it to be representative of the final game in a performance sense" right here http://www.hardocp.com/article.html?art=NDc0. Funny i dont see the 9800 or fx boards getting those frame rates now, whys that i wonder? Could it be the game wasn't finished? SURELY NOT??? :eek: :rolleyes:
Click to expand...

Nobody thinks going by a leaked beta is a good thing. But in the case of Half-life, that's all we have to go on. So if you were to try and guess the approximate performance of the X800 and 6800 in HL2 you could look at how they ran the beta and make some intelligent guesses. Didn't Valve themselves release benchmark numbers based on that exact same code and tell everyone that Nvidia cards suck in HL2?

In general, if you find yourself saying that these Doom3 benchmarks don't matter, or that HL2 beta benchmarks don't matter, then you're putting your blinders on. It ALL matters.
 
Arkayas said:
People are talking about how 'no one has gotten the code so no one could have optimzied for it' But it seems people are forgetting something.

Everyone except ID, and they did.. Carmak talked about having to put in a special code path for NV cards to make them work with Doom3 because their perf without it was crap. I want to see Benchmarks without NVOptimized code, having the engine optimized invalidates the game as a Benchmark of the cards.

If without the optimized code NV still beats out ATI then sure NV is faster in OpenGL (which is used by how many engines these days?) Anyone who didn't know that NV was faster in OpenGL must have lived under a rock, BUT when the engine has optimized paths for one card it becomes slanted. So is NV really faster in OpenGL anymore? or is it purely optimized engine code that makes NV faster... Hopefully we'll know sooner rather than later.
Click to expand...

This has been said about a hundred times already. THE NV OPTIMIZED CODE PATH WAS DROPPED MONTHS AGO. ATI AND NV ARE BOTH RUNNING ON THE SAME ARB2 CODEPATH.
 
Holy shet! I go to work see this benchmark release, come here and already friggen 28 pages. whoa! I read the 1st page and skipped to the end. I'll read the rest later on.

Anyways I can't say i'm total trusting in these benchmarks. Nvidia has been dipping in the game industy alot more than ATI. Hell, almost every new game that comes out has that nVidia video on startup. Once Doom3 actually releases and ATI gets a better look into the game I am certain we will see a tip in the scales.

So give it some time folks, new drivers are sure to make this crazy.
 
coffee33 said:
Does HL matter? You know your family? Yes them. Its more importain then them! hehe :D
Click to expand...


I can careless about HL2 to be honest..I never played the first one much...I been waiting for Doom3 forever
 
I just noticed that nivida broke the 60fps limit how is this so ? I mean isnt it limited to 60fps yet it showed 61 fps ? hmmm hehe just curious about this.
 
^eMpTy^ said:
Nobody thinks going by a leaked beta is a good thing. But in the case of Half-life, that's all we have to go on. So if you were to try and guess the approximate performance of the X800 and 6800 in HL2 you could look at how they ran the beta and make some intelligent guesses. Didn't Valve themselves release benchmark numbers based on that exact same code and tell everyone that Nvidia cards suck in HL2?

In general, if you find yourself saying that these Doom3 benchmarks don't matter, or that HL2 beta benchmarks don't matter, then you're putting your blinders on. It ALL matters.
Click to expand...


Leaked beta benchmarks are all we have to go on but it really makes no difference, hardocp benchmarked doom a year back in a state that carmack believed would show game performance, and even he was way off with his estimates. It was a controlled environment benchmark with the support of the developer not just some bunch of crap that was ripped off a valve pc with barely anything working in a game sense. I woud not put ANY stock whatsoever into the hl2 benches because it will not show what i would consider game performance in any sense of the word.
 
colin said:
You know the ATI guys/gals are working their asses off trying to get better OpenGL performance right now. Having recently bought my first ATI card (just a 9600xt), I hope they pull it off, and soon...
The debate doesn't effect me too much though, I don't have the money for any of those high-end cards...
Click to expand...



Just like they have been working their asses off for the past 3 years with better OGL support. But guess what? It still ain't there. I STILL have problems in Half-Life OGL. ATI, while having come a long ways, still have further to go with their drivers, and I have tested a lot of ATI cards when I have done reviews for GotApex.


LK
 
HL2 is going to be freaking awesome as well, a lot of games are going to be based on that engine. I would probably build another system w/ a x800pro just for HL2 if the 6800 sucks at it.
 
Impact9 said:
Holy shet! I go to work see this benchmark release, come here and already friggen 28 pages. whoa! I read the 1st page and skipped to the end. I'll read the rest later on.

Anyways I can't say i'm total trusting in these benchmarks. Nvidia has been dipping in the game industy alot more than ATI. Hell, almost every new game that comes out has that nVidia video on startup. Once Doom3 actually releases and ATI gets a better look into the game I am certain we will see a tip in the scales.

So give it some time folks, new drivers are sure to make this crazy.
Click to expand...

When you get done reading all 28 pages of this thread you'll know that this has nothing to do with optimized code and everything to do with Nvidia's superior hardware and OpenGL performance.

ATi has not been cheated. They shot themselves in the foot by sticking with an older core design and mostly not fixing thier OpenGL driver in more than 2 years of working with the r300 design.
 
creedAMD said:
HL2 is going to be freaking awesome as well, a lot of games are going to be based on that engine. I would probably build another system w/ a x800pro just for HL2 if the 6800 sucks at it.
Click to expand...
I hear you!
I would even buy a X800 if it meant a much better gaming experence for HL2! :)
 
^eMpTy^ said:
This has been said about a hundred times already. THE NV OPTIMIZED CODE PATH WAS DROPPED MONTHS AGO. ATI AND NV ARE BOTH RUNNING ON THE SAME ARB2 CODEPATH.
Click to expand...

A hundred times? Where? I have yet to see any release from ID saying that. If you have a link that would be more than helpful.
 
Chris_B said:
Leaked beta benchmarks are all we have to go on but it really makes no difference, hardocp benchmarked doom a year back in a state that carmack believed would show game performance, and even he was way off with his estimates. It was a controlled environment benchmark with the support of the developer not just some bunch of crap that was ripped off a valve pc with barely anything working in a game sense. I woud not put ANY stock whatsoever into the hl2 benches because it will not show what i would consider game performance in any sense of the word.
Click to expand...

I'm not saying we should consider them the end all be all...but that leaked beta was the same code that Valve tested on. Everyone thinks that ATi is gonna be so much faster in HL2 because of the benchmarks Valve released forever ago running on the NV30. The NV40 is a whole new core and it's going to perform right on par with ATi in HL2 just like any other game. And if someone runs benchmarks to confirm that I don't think you should take it with a grain of salt. Clearly all evidence showing ATi performing much higher than Nvidia in HL2 is outdated and running the beta (which is the same code Valve used to show ATi in the lead) on a 6800 proves that.
 
This has been predicted for years. Our pc's have become overgrown consoles. Titles exclusive to different videocards. This is only the beginning.
 
You must log in or register to reply here.
Back
Top