3dMark High polygon count?

[Gimps]

I've been looking over scores for the High polygon count(8 lights) on 3dMark, and it seems that a GeForce 2 MX400 can pull better scores than most GF3 Ti200's. I haven't compared the scores to the Ti500 yet, but most Ti200's pull right around 3.5-4.0 MTriangles/s. My GF2 MX400 just pulled 5.3 MTriangles/s, and it isn't overclocked:(. Can anyone explain this, or at least post some Ti200 scores and prove me wrong.

[NASsoccer]

i can't explain it, but for a reference my gf3 pushes 6.7 Mtriangles/s. I too would like to find out the answer to his post because my friend has a mx400 and it only pushes 3.6.

[Gimps]

As a reference to my stock benchmark with my GF2 try this compare URL:
» http://gamershq.madonion.com/compare2k1.shtml?2793828

I looked at the top scores for the GF3 Ti200 and my high polygon count is above all but 1 of them. Most of these compared systems are either overclocked XP systems or 2.5+GhZ P4 systems. The one that was above me was a P4 at 2583 MhZ and he was ahead by .7 MTriangles/s.

Any answers as to why, right now I'm thinking maybe its a RDRAM thing as I outscore AMD chipsets which are using DDR, but get beat by a P4(I'm guessing using RDRAM could be DDR though)?

[Pinky]

I'm wondering the same thing. You even beat my results (http://gamershq.madonion.com/compare2k1.shtml?2803615 ) ;).

I wonder if the newer chipsets use this feature or, like you suggested, it's pushed off to the cpu...

[Gimps]

Yeah its got me puzzled and the only variable that I can narrow it down to so far is the RAM. Would RDRAM really make that much of a difference over these overclocked DDR setups? I don't think so, but I can't see any other variable that is influencing this. I don't know why the GF3 Ti200 chipset would be getting this lower count especially when the GF3 is an improvement over my MX400.
BTW the 4600 is on the way!! Only 2 more weeks, but oh I'll be in Florida till March 10th!

*EDIT* Ok I went back to MadOnion and went through the Ti500 scores and I had to go down a dozen scores from the top of the list before I found an 8light high poly count higher than mine. Again it was a P4 system, but I'm wondering if maybe its in the GF3 chipset? All of these scores I checked were over 10k, but somehow they have lower poly counts than a MX400? Just makes no sense.

[rUfUnKy]

I don't know what gave me such a boost in my 8 light but after useing this tweak guide to the tea (except I also lowered my texture level to -3 in direct X) I made a huge jump!!!http://g256.com/guides/geforce/

Not Tweaked (Default clock) High Polygon Count (8 lights) 3.7 MTriangels/s ... http://gamershq.madonion.com/compare2k1.shtml?2209198

Tweaked with NVMAX (@ 271/501) High Polygon Count (8 lights) 6.5 MTriangels/s .. http://gamershq.madonion.com/compare2k1.shtml?2820559

[]-[itman]

hmmm i'm not too sure either. The only thing i could think of is that the newer cards have some setting like improved visual quality or something that slows them down. Also, look at your 1light ploygon count againts others. Your card gets smoked even by mine. That's weird that yours takes the lead in the 8light count.

[Gimps]

Oh I know the MX gets smoked in every category except the 8 light poly count. I don't know what it could be, I overclocked the card though and got that count up to 6.2 MTriangles/s but you'd still smoke me in every other category.

I guess its just one of those abnormalities of life that you don't like but you have to accept it anyway.

[SBeaver]

That wierd cuz my gf3-ti200 smokes absolutly every card I've ever compared it too, and by a large amount.