3dmark 2001

[Evil Gemini]

I only got 4039 on default.

I plan on getting an XP2400+ and a Leadtek Geforce4 Ti 4600 (if i can find that card any where) If not that card probably an FX 5600

Damn 3rd gen noob 11015!!!!






System Info Version: 2.2
Operating System: Microsoft Windows XP
Processor: AMD Duron @ 1.28 GHz
Graphics Card: NVIDIA GeForce4 MX 440 64MB DDR
Total Physical Memory: 512 MB
Free Physical Memory: 386.39 MB
Motherboard Manufacturer:
Motherboard Model: 8363/A-686A

[adamp2p]

I do not understand why you guys are running this DX8.1 benchmark with your DX9 software and (some of you) DX9 cards. I was thinking about running this benchmark too but I do not see the point.
Please explain.

Is there not another, newer 3dmark that more accurately measures 3d performance?

[DarthInsinuate]

apparently the latest build of 3DMark03 fixed the 'cheating' - not entirely sure about that though

[adamp2p]

...I am not trying to say that I am too good for this benchmark...I am just weary of installing/uninstalling too much as it harms my file system...

[Evil Gemini]

Its because 2001 doesnt demand as much as 2003 and 2003 makes you feel like you have a shit system when you get them low marks.

And also 2003 is bloody huge to download

[3RA1N1AC]

Originally posted by adamp2p@13 October 2003 - 15:43
I do not understand why you guys are running this DX8.1 benchmark with your DX9 software and (some of you) DX9 cards. I was thinking about running this benchmark too but I do not see the point.
Please explain.

Is there not another, newer 3dmark that more accurately measures 3d performance?

most of us don't have dx9 hardware. that's the point. bigdawgfoxx tried 3dmark 2003, couldn't run most of it, and got a terrible score... so i suggested that 3dmark 2001 would be a better test for his video card. 3dmark 2003 basically excludes any video card without at least directx 8.1 hardware support. so he ran 2001 and started a thread to ask for comparisons.

as for whether one or the other "accurately" measures 3D performance, i think you're mistaken in believing that either of them is an accurate measurement. synthetic benchmarks such as these may give you a rough measurement of performance in a generic "next-generation game"... but they have almost no firm relation to performance in real games, aside from Max Payne (the only game ever released that uses the same graphics engine as 3DMark). games are all programmed differently, and much of the performance comes down to programming and design skills on the part of the game developers. the people who created 3DMark may be better than some programmers at creating an engine that uses hardware resources efficiently, and they may be worse than others.

the fact that a geforce4mx is excluded from running most of the tests on 3dmark 2003 disqualifies it as an accurate indicator of how well the card can run today's games, since most current games do support that card. and then to say that the test measures performance in next-generation games... well, that is pure speculation on the part of the 3DMark creators, because they're not directly involved in producing the big upcoming games (personally, i'm not exactly waiting with baited breath for Max Payne 2 ).

if you want to know how well your video card is going to run Max Payne, or you just want something to brag about, run 3DMark tests. but 3DMark is not an accurate indicator of how well a video card will run Quake 3, Doom 3, Splinter Cell, Unreal 2, Half-Life 2, Tony Hawk 4, or any other games. especially not the ones that use OpenGL, which is completely different from Direct3D.

keeping in mind that running artificial tests is completely different from running actual games, there's nothing wrong with running 'em and comparing scores just for kicks. but the tests' relevance to the value of a computer is marginal, imho. the real test of a computer's value is whether it runs the programs that you need it to run at an acceptable speed, not the score it achieves in an artificial test.

[kurtsl0an]

3779.

Athlon XP 2600
512MB RAM
NVidia GeoForce MX 440 64MB

Good thing i'm not a gamer!

[adamp2p]

Originally posted by 3RA1N1AC@14 October 2003 - 01:19

keeping in mind that running artificial tests is completely different from running actual games, there's nothing wrong with running 'em and comparing scores just for kicks. but the tests' relevance to the value of a computer is marginal, imho.

Well said, mr brain.

Aquamark3 is a benchmark that tries to measure how your card will perform in future games as it generates a graphics score as well as a CPU score. If you ever read Anandtech (and somehow put up with a shitload of advertisements) you see that Aquamark3 is among the first benchmarks run as it is an aggressive benchmark that has some interesting features (in the full version) that enable screenshots at x frame and other nifty features.

You are correct about synthetic benchmarks, I am not going to argue there...but besides giving a kick, the benchmark is intended to accurately measure performance in a consistent enough way so that when comparing several results we can compare different systems and configurations, and performance in a numerical meausurement.

(my favorite thing about these benchmark companies is their online browser results, as it displays the specs of the computer that ran the test. That way us regulars on the board can see who is and who is not bluffing about their specs...not that it makes you better to have better specs, but why lie?)...

[noname12]

2692 (cries)

AMD Athlon XP 1800+
256 MB SDRAM
Nvidia Gforce 2 MX 400 64 MB DDR AGP 4x

[3RA1N1AC]

Originally posted by adamp2p@13 October 2003 - 16:36
(my favorite thing about these benchmark companies is their online browser results, as it displays the specs of the computer that ran the test. That way us regulars on the board can see who is and who is not bluffing about their specs...not that it makes you better to have better specs, but why lie?)...

guess i'll have to run the test again and provide a published score sometime, then. though if i were gonna bluff, i'd prolly claim to have better specs than i do. i'm guessing my score was pretty average, for a non-overclocked PC with those specs.

i don't think artificial bench tests are completely useless. even if it's impossible for these tests to be completely unbiased, they can give you some indication of one system's performance level against another. and if you compare to a bunch of other people with the same specs, it can help you figure out whether you've set your PC up correctly, or what kind of advantage you can expect from overclocking.

personally i'm just wary of using a bench test to predict performance in real world programs, because a good programmer can make a lean & mean program that produces great results on low spec hardware, and a bad programmer can easily create a hog that achieves very little on even the highest spec PCs.