3Dmark2003 benchmarks

[dolanenwindrift]

Was curious what numbers people were getting off their systems under 3Dmark2003.

I managed to squeeze 1009 3Dmarks from the following system:

Athlon XP 1800+ @1649 mhz
nVidia GeForce3 TI200 Core @ 200 mhz, memory clock 450 mhz
ECS K7S5A
512 MB PC2100


Dol

[Betty]

927...system is as follows:

shuttle ak35
1.5 gig athlon
256 mb ddr ram
radeon 9000 128 mb ddr
sb value live sound
15 gig quantum hd (severely needs upgrading)
OS is winXP pro

All stock speeds, nothing is overclocked...all latest drivers ect. installed.

[robbie]

Abit KD7 with a 1900+
PNY GF4 4400
512 of PC3200 mem set to Ultra

1496 was my score. I guess I really need to tweek my system.
Rob

[marcofra2000]

Originally posted by robbie
Abit KD7 with a 1900+
PNY GF4 4400
512 of PC3200 mem set to Ultra

1496 was my score. I guess I really need to tweek my system.
Rob

Keep on tweaking dude. With my setup I got like 47XX I think almost the 5000 barrior with not much tweaking and overclocking.

[BrianW]

I heard that this benchmark is slighted towards ATI? Any takers?

[dolanenwindrift]

The benchmark, according to Futuremark, is GPU neutral. Apparently the Radeon scores much better without any specific optimization than the nVidia cards, so nVidia is raising a bit of a stink. There is a good article on Tom's Hardware about this subject. I'm sure Anandtech has some info as well, but I don't visit his page.

Dol

[marcofra2000]

Originally posted by BrianW
I heard that this benchmark is slighted towards ATI? Any takers?

I don't think so, but 3dmark03 supposedly (spelling?) uses pixel shader 1.4 and nvidia card can't do that like the geforce 4 and under. So the comparing is not that equal and ATI gets the advantage on this which their cards can render 1.4 pixel shader. Nvidia says that no games is/are using going ps 1.4 which nvidia claims this benchmark is not as good as 3dmark01.

[Liquid3D]

My system seems to really like 3DMark2003, because I'm only getting 14477, in 3DMark2001SE? Here's my sys spec's: AXDA2400 - 2254MHz (13.5x166) AIUCB 1.9V, Epox 8K9A (VLink 8X Support disabled resolving VIA 4.45 driver conflict w/R300), 1024MB OCZ EL-PC3700, Tyan Tachyon G9700 Pro (340/320), Antec TruPwr 480W, WinXP SP1, Cat 3.1, Hyperion 4.45. AGP 8X. Here's a pic.

 

[marcofra2000]

Nice score. Can we see you breaking 5000 probably?  

[dolanenwindrift]

Awesome man! Looking at a new mobo/CPU soon, hopefully I will see some better numbers myself.

Dol

[Meridian]

2160

Leadtek Ti4600 @ 320/744
XP2700+ at 178 x 13.5 (2416)
Epox 8K9A2+
512MB Corsair XMS3200


Now now way is a 9700 twice as fast as this card (I've benched them together for a review and I'd say 20% max at the resolution used for this bench), so I have to agree - ATi bias.

M

[romper77]

this is with 9700pro which at this time is being rma ed (never mind)
xp-2400@2431

[marcofra2000]

WOW, all you guys have realy good scores. I'm going to try to break 5000 this weekend since I don't have time during the week (damn and is hardly monday )

[SteveI]

Originally posted by Meridian
...Now now way is a 9700 twice as fast as this card (I've benched them together for a review and I'd say 20% max at the resolution used for this bench), so I have to agree - ATi bias.
M

Dude... guys.. get real.
3d2k3 is a DX9 based test... Futuremark is pushing it with the logic that this is supposed to test current hw for FUTURE capabilities with games that haven't even been developed yet.

This is not an ATI bias at all... The reason SOME ATI cards do well is that SOME ATI cards are DX9 capable

[marcofra2000]

Originally posted by SteveI


Dude... guys.. get real.
3d2k3 is a DX9 based test... Futuremark is pushing it with the logic that this is supposed to test current hw for FUTURE capabilities with games that haven't even been developed yet.

This is not an ATI bias at all... The reason SOME ATI cards do well is that SOME ATI cards are DX9 capable

I couldn't say it any better

[Liquid3D]

Well I have a brand-new 8RDA sitting on the floor next to me in a box, so YES I think I can beat it!!!! BUT I went to repeat the overclock and benchmark, this afternoon and couldn't do anything as the 3DMark2003, continually crashed to Windows. No matter what I did (beyond running it at default speeds) but what's the point after you've hit a certain bench? I thought there would be headroom with this card, and I've made NO system changes since the original benchmark? I'm thinking the card is a very poor performer, recent 3DMark2003 scores not withstanding, as it's only scoring 14K in 3Dmark2001SE?

 The Tyan is unable to overclock above 350/325, and it's driving me mad! I'm quite upset having spent $370 on a card designed from the ground up for "performance" with an aliminum machined heatsink the like of which can only be found on Leadtek cards. It's very dissapointing they send a sample to [H]ardOCP which attains 385/340 and I've yet to hear a Tyan Tachyon owner proclaim speeds above 360/340. I'm most definately RMA'ing this card, and iof they give me the "it meets it's default clock speed requirements" speech, I'll retort with the "Your going to be eviscerated in the Hardware Forums of the Planet" speech!

I understand every new product must undergo growing pains, hence the term "revision; a" however; they could at least send out cards with close the overclocking prowess of Sapphires' and even Powercolor's at $100 less!!!

[marcofra2000]

If you think the Rad 9700 Pro is not so good as you expected to break 5000, then wait for the next driver release 3.2. Supposely they will rock incrementing fps.

[dolanenwindrift]

Got my system to post a 1055, ran the Geforce3 TI 200 at 210/475 mhz.


Dol

[Redrum]

Its a tough benchmark and can make the best cards sweat. I made it through with a score of 5065.

Computer Specs

[marcofra2000]

Great score.

My 3dmark03 score is now 4915. Great score from what I have on my Sig down there. I'm going to try to break 5000 during the day without overclocking-just tweaking. If I can't, then I'll overclock.