3dMark 2001 Nature test

[Cat13]

I am getting around 10,000 with my current setup in 3dMark with all of my fps's equal to the ones I compare to, except the nature test. I seem to be stuck in the 40's on this test. Does anyone have a suggestion to help me get this particular score higher using the default settings for 3dmark and DX. I don't want to sacrafice image quality for higher scores. Here is the system in question: rig I am overclocked using a fsb of 147, no multiplier or voltage tweaks. The video card is stock.

Thanks for the inputs

 

[AnAndAustin]

Because the Nature test uses the specific DX8 hw, you will find maximum perf will come from o/c the gfx card, in fact generally o/c the gfx card will boost most PCs gaming perf much more than CPU/mobo o/c.

10000 is a very nice score, esp without compromising quality. TI4400 is at 275/550 at default and you should get around 300/620. Obviously you want to use small increases until you find your particular cards limit. Let us know what diff this makes for you, best of luck!

 

[Cat13]

Well I messed around with overclocking my video card and raised my score up from 10043 to 10712. In the end I bumped it up to 300/650, stock 4600 speed. everything ran fine, no artifacts, actually looked really good. I am running 25C case/40C cpu so heat is not really a problem, but I would like to know what the card is running. I will probaly put a temp probe on my card.

Overclocking the video card helped, as my fps's in nature went up about 10, guess thats all I will be able to get for now. Any other pointers?

Thanks!

 

[OneOfTheseDays]

here's what i found out. You need to use a program like RivaTuner that can adjust your LOD settings. By setting the LOD very high, it raised my nature score from mid 40's to close to 60 fps.

 

[Cat13]

but doesn't that take away from image quality?

 

[Krk3561]

Whats LOD?

 

[Cat13]

I ran riva tuner and used the database for the 29.80's. Once I got the lod (level of display) slider to show 15 instead of 3 I did some benches. Obviously if I went up, the fps would increase with a dramatic lose in image quality. When I went "negative" I also got a performance boost, not as much though. I did not notice a difference between say -8 and -15, and image quality was for the most part the same as 0. Guess I can mess with rtuner a litle more and see what it can do.

thanks

 

[Barrei]

Chage from best image to best performance in Direct 3D settings , should up your score by 200 - 400 points.

 

[OneOfTheseDays]

well you wanted to know how to raise your nature score and that is how these guys with higher scores are doing it. They muck around with the LOD Bias settings and sure it looks ugly as hell, but the score rises dramatically.

 

[Barrei]

I never changed it for my posted scores , every time I read one of your responses they always have a BITTER edge to them , smile once in a while you might actually begin to enjoy life a LITTLE bit.

 

[Goi]

Geez...do you actually derive immense satisfaction from a higher 3DMark score? I mean, comon, lowering the LOD and such to gain a few hundred more points on 3DMark? I can understand if its a game, and performance counts, and correct me if I'm wrong, but isn't 3DMark just a benchmark? I dunno, maybe its just me...but whatever rocks your boat man...

Cheers!

 

[lRageATMl]

maybe i've been assuming too much, but with everyone's posted benchmarks I assumed that they left ALL the settings alone (besides overclocking) so we can set some kind of standard in which we can compare our benchmarks. sure I can reduce the quality completly and run at the smallest resolution, but come on...do I wan to? or do you really want to?

 

[Barrei]

I agree all posted benchmarks should be run without tinkering , and Goi look in the mirror it is just you.

 

[AnAndAustin]

I think 3Dmark is very useful for giving a general idea of what the perf of your PC should be. If it falls short then it can point to a couple of simple o/c, BIOS or driver tweaks (not lowering quality) which can boost you a good 10% or so. I don't think anybody thinks attaining the most marks is the important thing, just checking that their system gives about what it shouold be.

Another great advantage of 3Dmark is to consider what perf is possible through different upgrades. Of course things will always vary to a degree depending upon the particular game, but as a general guide goes 3Dmark is pretty good. Whenever I quote or read marks for a given configuration I never (where possible) quote the highest scores, I scroll through the selection a page or two in and look for where the scores calm down a little, this gives the best idea of what a particular configuration should achieve.

 

[OneOfTheseDays]

yea i too disagree with 3Dmark scores because i believe they should be run with the exact same settings on all cards. And it really sucks taht people are tweaking hteir LOD's just to gain points and get higher scores. That's why if you see a guy with close to 60 FPS on nature, you can almost guarantee that he is using the LOD tweak as that significantly boosts nature scores.

 

[Cat13]

This is exactly why I am having so many problems getting a good idea if my system is running to it's max potential. I try to use 3dmark to guage it against other simliar setups but so many are inflated due to tweaks that would not be used during everyday gaming or what not. I want a higher score to let me know my system is running good, but I will not be satisfied if I don't enjoy looking at the screen because the graphics look like my 266 with onboard video. Image quality is what I want, but it's nice to have an idea of how well you rig is running.

Thanks for all the help!

 

[AnAndAustin]

Originally posted by: AnAndAustin
I think 3Dmark is very useful for giving a general idea of what the perf of your PC should be. If it falls short then it can point to a couple of simple o/c, BIOS or driver tweaks (not lowering quality) which can boost you a good 10% or so. I don't think anybody thinks attaining the most marks is the important thing, just checking that their system gives about what it shouold be.

Another great advantage of 3Dmark is to consider what perf is possible through different upgrades. Of course things will always vary to a degree depending upon the particular game, but as a general guide goes 3Dmark is pretty good. Whenever I quote or read marks for a given configuration I never (where possible) quote the highest scores, I scroll through the selection a page or two in and look for where the scores calm down a little, this gives the best idea of what a particular configuration should achieve.

Even with identical configs you can still find o/c'ing varies quite a lot. Not only that but 3Dmark version, BIOS settings and driver versions can also make a fair difference too. Even if the top results for a given config are genuine and don't destroy image quality it is still unlikely that the vast majority of people with very similar configs can achieve it, just remember a KT133A is about 10% slower than KT266A even with all other components, settings and drivers being identical.

 

[Cat13]

Thats why I think 3dmark needs to take variables like that into consideration so we can compare "like" systems to get the most accurate comparison.