Correlation Between 3dMark 2003 & HL2?

N

Nutzack

Guest
hi,

if this topic has already been brought up, please flame me.

but for the sake of argument: i check in at just < 5000 marks on 2003...

i have an 1800+ right now, and 512 2100 :[ RAM, with a 9700 pro...

do want to run single player with most bells and whistles, a little aa and af...i wonder of this is possible...thx in advance for your input.

Nutzack
 
1)A way to prevent flaming: search the forums.
2) This should be posted in the Hardware forum
Edit: Don't worry, I won't flame you:cheers:
 
Yea, you should be able to run with max settings even with AA and AF turned on, but don't take my word for it... wait for the benchmark.
 
Why wouldnt this belong in the hardware forum?
 
well 3dmark03 tells you weather or not you have a good dx9 video card, which you do. you might want to upgrade your processor, but 3dmark03 won't tell you that becuase it's almsot entirly a vidoe card benchmark.
 
Originally posted by simmo
AA decreases performance right ?
Click to expand...

not that much. I have a ti 4200 and i put on 4x AA and 4x AF and i only lost like 5 fps in DOD, i get like 40-90 in dod, but 99.9 in all other hl mods. I lost like 1 fps in americas army with AA, but af really did take a lot of fps in that game. AA shouldnt take much in anything so i think you should turn that on.
 
eeeh LOL?
I have an amd athlon XP 1800+, 256 pc-2100 ddr, and ati radeon 9700 pro, and my 3dmark03 score is 4338, why so much difference between us, just for the mb ram?
 
Originally posted by Xenome
eeeh LOL?
I have an amd athlon XP 1800+, 256 pc-2100 ddr, and ati radeon 9700 pro, and my 3dmark03 score is 4338, why so much difference between us, just for the mb ram?
Click to expand...

Hmmm, I've got an 1800+, a gig of ddr pc2700 and a 9700 Pro and my score was only 4160... ;(

Anyone have any tips or suggestions??? PLEASE!!!

I always shut down all programs (from the taskbar and everything) and run it on a fresh boot. I want my money's worth, dammit!

EDIT: What OS are you guys running? I've got WinXP Pro. Would the GUI features of XP slow it down much? I don't turn them off...
 
Ive got windows 2000 pro service pack 4, and in the driver settings (catalyst 3.6)everything on high performance; aa off, anistropic filtering off, enz enz everything on low. In my last benchmark i got 4327 btw :p
 
Originally posted by Xenome
Ive got windows 2000 pro service pack 4, and in the driver settings (catalyst 3.6)everything on high performance; aa off, anistropic filtering off, enz enz everything on low. In my last benchmark i got 4327 btw :p
Click to expand...

Hmmm, maybe I'll have to do some tweaking tonight! :cheese:
 
The only thing to tweak is your video card. i belive it was [H]ard OCP that ran 3dmark 03 on a p3 400mhz machine with a 9700 pro, and only had a 100 or so fewer #D03marks than a 2.8Ghz p4 with the same video card. All 3DMark03 does is test your video card.

He mgiht of gotten that score becuase he had lower setting (ie 800x600 instead of 1024x768, 16but color instead of 32bit) things like that.
 
Originally posted by SidewinderX143
The only thing to tweak is your video card. i belive it was [H]ard OCP that ran 3dmark 03 on a p3 400mhz machine with a 9700 pro, and only had a 100 or so fewer #D03marks than a 2.8Ghz p4 with the same video card. All 3DMark03 does is test your video card.

He mgiht of gotten that score becuase he had lower setting (ie 800x600 instead of 1024x768, 16but color instead of 32bit) things like that.
Click to expand...

Really? So the score doesn't take into account what resolution and color setting you're running at? I had always figured that if you tuned down the resolution it would lower your score accordingly. Hmmm, and my respect for 3dMark goes down again...

c'mon HL2 Benchmark!!!!! :cheers:
 
i oc my card with ragepro...not by much: core = 350, memory clock = 330...

i thought that since there is a cpu test in there, it would reflect how well my machine would do on the higer end games...anyway, thx 4 the input guys.
 
I'm presuming a GF3 isn't a DX9 compliant card?
 
No i was sayignt hat it deos take into accoutn resolution. the lower res the more 3dmarks.

the cpu test has a negligable impact on your score.
my p4 1.8GHZ machine got 27 3d03 marks less than an amd xp 3000+ did with the smae video card.
 
Well then i guess that P4 pwns the 3000, which means it pwns the 3200+!
























hehe:p
 
Originally posted by SidewinderX143
No i was sayignt hat it deos take into accoutn resolution. the lower res the more 3dmarks.
Click to expand...

I think I understood what you were saying before. Do you mean that it's possible to artificially inflate your "score" by lowering your resolution? For instance, if I run on my machine at 1024 resolution and get a score of 4500 and then run with a resolution of 800, would I get a higher score the second time?

I would expect the scoring mechanism to handicap a card running at lower resolution and lower the score accordingly, as the card hasn't gotten faster, just the framerate.

Does my question make sense?
 
i run 3dmark 2003 at default settings...i bet if you could lower the 1024 default, you could get better scores...less work on the card...however, wtf would they put a cpu test in there if it accounts for nothing?:flame:
 
3DMark 2k3 is pretty much useless as a benchmark. I don't trust it at all, not after the 'incidents'
 
Originally posted by [[LuCkY]]
3DMark 2k3 is pretty much useless as a benchmark. I don't trust it at all, not after the 'incidents'
Click to expand...

I'll agree with that, but it's still a nice way of comparing two boxes...especially if they have similar specs.

I mean, just from checking this thread, I've realized that my system's 9700 pro is a fair clip behind some other people's 9700 pros. Some tweaking is probably needed.

And for that, I think that synthetic benchmarks do a decent job. Although, a real world benchmark could also accomplish the same thing. But the 3DMark scores do wrap it up easily for that. It would be cool if Valve established some sort of scoring system in their HL2 Benchmark. And maybe using steam as a place to compare your rig to others. Then I'd gladly leave 3DMark in the dust!
 
no, the CPU test account for soemthing, but very, very little.

It's a nice way of comparing two video cards.
 
I've got a:
P4 2.8 GHZ
1 GB DDR (3200)
9800 pro 256 MB

And only got 2400 points WTF!

Any suggestions? I ran it with default setting by the way.
 
Originally posted by mattjackson
I've got a:
P4 2.8 GHZ
1 GB DDR (3200)
9800 pro 256 MB

And only got 2400 points WTF!

Any suggestions? I ran it with default setting by the way.
Click to expand...

i wouldnt worry much about 3dmark.

Thats an awesome system you got yourself. You shouldnt worry about what benchmarks say for a system like that. If you love it and it is f*ing great in games and whatever you do then your good.
 
I Know it doesn't matter what the benchmark score is, its the game performance that counts. I am just curious as to why the 3dmark score was so low. At some points in the test it slowed down to about 3 or 4 fps.

I've only had the computer a few days so haven't had chance to test much out on it yet. I've ran unreal 2, tron 2, GTA vice city and a few others all on max resolution and all other max visual settings (including 6X FSAA) and haven't experienced a hint of slowdown so I'm sure there's nothing to worry about.

Even so any suggestions anyone might have as to why the score is so low will be appreiciated.

(I have the latestest drivers installed)

Cheers
 
in your disdpaly settings (advanced) settings, do u have any antistropic filtering or anti aliasing enabled? the recommended settings are default, which means clicking the bar to quality (disabling aa and af).

your system is great, that card rocks and will for a long time.
 
I think I might of had all the settings set to max.

If these setting are disabled in control panel does that mean that when playing games, say AA will be disabled.

Thanks for your help
 
we ahve the same interface in our display: i just have opengl and direct set to quality...aa and af are turned off with that default setting...lol i tried to run 2003 with my card jacked way up and got like 2000 points, was like wtf...

btw, on a side-note, will HL2 have opengl features?
 
Well, I went home last night and dropped all the settings to performance and my score was 4156. An actual drop of about 4 points (hey, at least it's consistent). Once again, my system is 1800+, 1 gig of PC2700 DDR, and a 9700 Pro. When I see other people running almost the exact same system and getting well over 5000, I'm a little concerned.

Obviously, I don't care much about the 3DMark score by itself, but I haven't been as ecstatic with the actual gameplay performance of my machine with the 9700 Pro as I would have expected, so I'm wondering if this could point to some other issue in my box.

My m/b is decent (Gigabyte GA-7VAXP). I had all non-essential programs shut down and all special WinXP GUI stuff shut off. Here are my results: http://service.futuremark.com/compare?2k3=1156765

Can anyone check these out and maybe tell me if I'm doing something wrong? I'm pretty computer literate and nothing I can think of, short of bad hardware, would explain the short falls.

Thanks in advance!
 
turn your AGP to 4x or 8x. it says 0x.

other people's resolution might be lower than yours
 
Originally posted by trantjd
Well, I went home last night and dropped all the settings to performance and my score was 4156. An actual drop of about 4 points (hey, at least it's consistent). Once again, my system is 1800+, 1 gig of PC2700 DDR, and a 9700 Pro. When I see other people running almost the exact same system and getting well over 5000, I'm a little concerned.

Obviously, I don't care much about the 3DMark score by itself, but I haven't been as ecstatic with the actual gameplay performance of my machine with the 9700 Pro as I would have expected, so I'm wondering if this could point to some other issue in my box.

My m/b is decent (Gigabyte GA-7VAXP). I had all non-essential programs shut down and all special WinXP GUI stuff shut off. Here are my results: http://service.futuremark.com/compare?2k3=1156765

Can anyone check these out and maybe tell me if I'm doing something wrong? I'm pretty computer literate and nothing I can think of, short of bad hardware, would explain the short falls.

Thanks in advance!
Click to expand...

Nutzack overclocked his radeon9700pro abit, i think that's the difference.
I have almost the same system as you (except for ddr: i have 256 pc-2100 ddr)
I get 4322 points. That's not much different then yours.
Btw, have you set BOTH your openGL and direct3d settings to high performance?

PS. At your results i saw that your don't have the latest patch for 3dmark03 yet : 330 or something. Download it!
 
Originally posted by Xenome
Nutzack overclocked his radeon9700pro abit, i think that's the difference.
I have almost the same system as you (except for ddr: i have 256 pc-2100 ddr)
I get 4322 points. That's not much different then yours.
Btw, have you set BOTH your openGL and direct3d settings to high performance?

PS. At your results i saw that your don't have the latest patch for 3dmark03 yet : 330 or something. Download it!
Click to expand...

Yeah, I downloaded that update last night, but I haven't run it yet. I'll get that running tonight. As for the OpenGL settings, yeah, I got both DX and OpenGL.

Do you have a comparison link for your results so that I can check them out after I run tonight? If you don't, you just need to publish the results and it'll give you a link.

Thanks!
 
Well I got a 5787 on 3DMark03 stock w/o overclock. I'm thinking that anything above 5000 or so should HL2 perfectly with max settings, 4600-4999 fair with max settings, and anything below 4599 you should probably not use max settings. But these are just predictions. That and benchmarks don't mean everything.
 
Originally posted by Miasma
Well I got a 5787 on 3DMark03 stock w/o overclock. I'm thinking that anything above 5000 or so should HL2 perfectly with max settings, 4600-4999 fair with max settings, and anything below 4599 you should probably not use max settings. But these are just predictions. That and benchmarks don't mean everything.
Click to expand...

Hey Miasma, have you published that result on futuremark.com? I'd really like to take a look at your specs and compare it to mine. Can you post a link to the published results?

Thanks!:cheese:
 
Originally posted by trantjd
Yeah, I downloaded that update last night, but I haven't run it yet. I'll get that running tonight. As for the OpenGL settings, yeah, I got both DX and OpenGL.

Do you have a comparison link for your results so that I can check them out after I run tonight? If you don't, you just need to publish the results and it'll give you a link.

Thanks!
Click to expand...

3dmark03 result :

http://service.futuremark.com/compare?2k3=1158288
 
You must log in or register to reply here.
Back
Top