Official nVidia 3DMARK03 thread!

Like I said @ FM forums, I'm going to keep benching 3dmark2k1 for a long time still, 3dmark03 isn't fair and pretty much sucks, I'm just messing with it

 

1.1 and 1.3 are pratically the same. Theres not a big difference like there is with 1.4 (over twice the instruction slots and 6 textures/pass instead of 4) so they probably didn't bother with a 1.3 path as it wouldn't have saved any effort over using 1.1. How about doom III for a (albeit future) game that uses the equivalent of PS1.4

 

Yes! Can you imagine that it actually doesn't work on Matrox M3D and VooDoo1 cards?
Doom III will be unfair too.
People don't get one thing - 3Dmark is always ahead of time. GF4 is outdated by now (compared to 9700 which is one generation ahead, and NV doesn't have a solution to compete). It is like I say - I have best KYRO card (KYRO II) and it doesn't work. It is unfair.

 

I had the problem where the thing would stop all together- (not freeze- just stop in the cpu test) with all the drivers, but I was using a third party refresh fix and a program called Rivatuner.
then I reinstalled the driver and DID NOT use a third party refresh rate fix, ran the test again and it worked fine-
I installed the 42.01's and they also worked fine. I installed the 42.68's and I get a big band from half way up the screen to 3/4 up the screen from the bottom that is dark and only shows some of the textures (but it still scored well)- this happened with multiple re-installations of the driver. I don't get any of the errors you talk about though.
The 42.01's work fine, but don't score well. I'm gonna try the 42.86's again and see if I get that weird band in the middle of some of the tests.
I don't think you're the only person having unexplained problems with these drivers or the new benchmark.

 

There is something seriously wrong with this new benchmark !

I have a GFTi4600, ok, I know not as good as the 9700Pro, but it runs any game that I

throw at it without any problems. Running this new test was a joke ! I was getting

2 or 3 fps at some stage. :wtf:

I smell a rat somewhere, and NO its not sour grapes

My score :hmm:

 

Couldn't agree more Tosh

 

Guys, try to remember that this benchmark is quite a ways ahead of the current hardware. How many of you remember when 3DMark99 came out... who remembers trying to run that on the Rage Pro, and the Rage128... for those that dont remember, it was a friggen slide show. It wasnt until nearly a year later that the frame rates started to rise! 3DMark2001 was 2 years old! it no longer represented the leaps in technology, and it really was no longer a good benchmark because it never stressed the hardware. I remember getting a few thousand points on 3DMark2001, and feeling like I had accomplished something. Now all of you can shut up, and run this benchmark as what it was intended to be... A BENCHMARK, it is not life! Although this does represent how well your hardware will run the games that will be out in 4 years.... LOL.... now... if you want to bitch, go to www.futuremark.com, register in the forum, and bitch to your hearts content. Running a 2 year old benchmark on hardware that is brand new is counter productive. In two years the hardware will once again be outrunning the software, and you will then again have 10,000+ mark scores... Use your heads people this is not an old benchmark with new hardware, it is a new benchmark with old hardware. When it comes down to gaming... if it plays, play it..! Now... I am done with my rant... Hell guys, I am actually quite pleased that my year and a half old card did this well. Now... anyone else want to gripe? **places paw over the ban button**

 

Sorry WyretheWolf but I have to disaggree with you on this one......... if that is still

allowed on this forum? :hmm: If we don't like the test, why the **** are we not

allowed to say so ? :wtf: This has nothing to do with DriverHeaven so why are you

threatening to ban people who want to gripe about it ?

 

I see your point, but you are completely missing the facts that futuremark has already commented on this. You are running a benchmark that is new... with old technology! And yet you are surprised that your hardware is not running the benchmark like it did with a 2 year old benchmark?!

 

Wolf's right guys..
Running this current generation benchmark program on old generation
hardware will always result in low scores, no matter how much you
run the tests over and over..

here's a question:
since 3dmark2k3 just came out, does that mean all the games that you
current play become chuck-slow? if my computer serves any indication,
NO.

B!tching and ranting about how your old card doesn't do well on a new
benchmark is useless.

My card only gets ~1400.. damn.
but the games that I play run hell smooth. and that's what counts.

Futuremark already spoke about this.. keep to 3DMark01 for our cards,
unless you have a DX9 ready card.

 

And since I can't edit my post, go play with this result for a while:

http://service.futuremark.com/compare?2k1=552432

 

i got 1263

 

Well.

I got a 3dmark03 score of.... 806 3d marks


On my little bro's PC
Compaq PC

Intel P4 2Ghz CPU

Compaq mobo with sound/N.i.c.

512 MB ram

Nvidia GeForce 3 200Ti 64MB AGP 4X

56kbps modem

Firewire card

I must say that as a benchmark its a very disapointing one and does not really represent the leaps made by the "pixel shader" cards.

There are many versions of "pixel shaders" and it would have made more sense for Futuremark to take their time researching "pixel shaders versions" and made the 3dmark03 a bigger benchmark program (over 500MB) so that it would include all the needed code so that the benchmark would be more demoratic, instead of looking like Florida's "punch a hole" elections.

Further, the XBox as covered in http://www.anandtech.com/showdoc.html?i=1561&p=1 and tom's hardware page.

Although its a nonupradable console that is somewhat obsolete compared to the latest cpus and vidcards, Quake 4, DooM III, Perfect Dark 2 for XBox and Halo 2 will be coming to the console soon.

Although its a different platform than the PC if these games come out and make 3dmark03 look outdated as Real games often do, its just going to reveal that Futuremark made a boo boo and they need to fix it with a patch soon.

The reason I say this is because after this benchmark came out and it did not seem to fully use Dx8 or Dx9 at all, and when Quake 4 and DooM 3 come out and run like butter on my bro's system (as the leaked DIII did) it will further show that either Direct X has a flaw compared to OGL or that Futuremark is helping ATI to stay king or become king for years to come.

I also have an ATI Fire GL 8800 128MB AGP 4X card that I do not use for games due to the low quality ATI drivers, but as soon as ATI releases a REAL new driver then I will test that card to see what kind of score the ATI R200 can get.

Also I wanted to add that it seems kind of unfair given the current state of the job market and the Economy to bring out a flawed benchmark that will more than likely make people buy new hardware so that they can get "higher framerates and scores"

 

114 with my Geforce2MX 400

 

well, amuro..
if you think about it -- any game that comes out for
consoles are always degraded in image quality.. not
to mention that they are running at a much lower
resolution than PCs do..

if you change 3dmark to be the same quality as it
would be on an xbox, it would sure run pretty
quickly..

but i'm thinking that doom3 will be quite fine on most
systems, even if you have to lower graphical settings
or resolutions.. there was a time when i wasn't able
to play quake2 in anything but 640x480..

we won't always have a super-fast card.. with the way
things are going, when we get a card -- it will always
be slower within 6 months.. but to me, a vidcard will
always last pretty long..

it'll still be a while for those games that you mentioned
to come out.. and by then, you'll probably have upgraded
something in your computer.. be it more ram, a better
cpu or a better vidcard..

 

115 http://service.futuremark.com/servl...rojectdetails&projectType=10&projectId=316111

 

1547 here with geforce 4600
however xp 1700 capping me

 

But you guys .... running the test seveal time get you huge diffrences.... the fact the the games put in the bencmark are not programs liked a game. I think is was toms hardware that was talking about it skinning objects in a sene 32 times per fame or something.. the software uses diffrent methods of rendering based on what card you have in the tests... I mean u and i cannot even begin to compare our bechmarks unless we are running the exact same cpu/mobo/videocard setup other wise ther not comparable... I am going to stick with 2001se for a while... dont get me wrong I still use them both I just dont rely on the 2003 score as realistic..... I am doing my own figuring right now... to much flac flying around about it to find sorces to 100% rely on..
3dmark2003 needs more tests kind of like those missing that were in 2001se

verry funny !!! doin't exadurate 3DMark2001 SE (build 300) was released in February 2002 .... that makes it just over 1 year old

 

Flak-Cannon to the face! (Futuremark's, that is)

 

3DMark03

OPS