3D Mark 2001 SE Today?

Last time we were given word from MadOnion, it's a PS 1.4 test, but only a "functionality" test. It does not contribute to the 3D Mark score.

In other words, there is a tech demo in 3D Mark 2001 SE for Radeon 8500 cards. But the fact that GeForce4 can't "fully" run it, won't negatively impact 3D Mark scores for GeForce4.

What we don't know, is if MadOnion did anything with the new Pixel Shader test that could help us analyze the real-world difference between capabilities of PS1.1-1.3 and 1.4. (Does the new test support several levels of PixelShader versions?)
 
Meaning:

nVidia will pressure MadOnion to include a PS 1.3 (or 1.1) version of the new PS1.4 test, that shows GeForce3/4 doing "the same thing" as a Pixel Shader 1.4 card...even if it's not rendering exactly the "same thing." ;)

I don't really think that will happen...but if it does, then there would seriously be something "wrong."

Basically, we're all just taking pot-shots at the apparent "closeness" that nVidia and MadOnion seem to share. ;)
 
Truly a show of favortism towards one company yet supposedly a unbiased benchmark. Madonion lost my respect with the peformance analyzer, people running a Kyro would be told their card is too slow and upgrade to a Geforce 2 MX...blehhh

I personally don't care what they do I won't be downloading it, If I want to see what Pixel Shader 1.4 does I'll use the ATI Nature demo....pathetic.

Truly a opportunity to compare speed and IQ with different versions of Pixel Shaders which I for one was looking forward too...oh well I'll rely on game benchmarks in the future.
 
just in case someone is interested ;)my GF2MX/1.2 Ghz Athlon has 2417 Points.

but the new 3dmark was much slower then the old one (the score isnt very different, but just looking at it shows that the old one was much more fluent). maybe NV needs to tweak their drivers a little bit
 
I said:

nVidia will pressure MadOnion to include a PS 1.3 (or 1.1) version of the new PS1.4 test, that shows GeForce3/4 doing "the same thing" as a Pixel Shader 1.4 card...even if it's not rendering exactly the "same thing."

I don't really think that will happen...but if it does, then there would seriously be something "wrong."
Click to expand...

Looks like I spoke a few minutes too soon. Right from Mad-Onion's FAQ, specifically related to the new "Advanced" pixel shading test:

Q: I have an ATI Radeon 8500, which should draw the water surface in the Advanced Pixel Shader test in a single pass, compared to my friend's system with DX8 hardware that should draw it in two passes. Still I don't see much performance difference. Shouldn't my system be twice as fast?

A: The Advance Pixel Shader test is what we call a Feature Test, which means that we, above all, want to present some new technology. It was decided that a fall-back was to be included in addition to the 1.4 pixel shader, since the same looking effect can be achieved using pixel shader 1.0 hardware. These two different modes of that same test work a bit differently and should, therefore, not be directly compared. Both modes could be optimized to show more performance either way, but now the test is just optimized for maximum compatibility. Vertex shader performance also affects the score, somewhat, due to this compatibility optimization.
Click to expand...

I'll be damned if MadOnion didn't do EXACTLY what I thought they wouldn't do because it's just "wrong." :cry:
 
I'm not sure I see the fallabck as 'wrong'; a tad daft to say they shouldn't be directly comparable.

Who wants to be the GF4 Ti's end up being faster than 8500 doing it though...
 
Umh..now the lobby test works fine on my Kyro 2, but the polygon troughput test is completely fucked up. Missing polygons strips, flashing, deadly slow.
The same happens with the vertex shaders test :cry:

ciao,
Marco
 
The issue that if it is truly a PS 1.4 test, it should be written in PS 1.4 code. Fall-backs are for games. If they wanted to show something that can be done just as good in PS 1.0, then it is not a true PS 1.4 test. They should have created something either:

1 - Only can be done via PS 1.4
or
2 - Can only be done via PS 1.1 / 1.3 via Multi-pass.

I would say the same thing for PS 1.3 as well. Do apples to apples.

<font size=-1>[ This Message was edited by: nooneyouknow on 2002-02-12 22:50 ]</font>
 
Hi Dave,

I'm not sure I see the fallabck as 'wrong'; a tad daft to say they shouldn't be directly comparable.
Click to expand...

If the scenes generated by the "fall-back" and "full 1.4 PS" are different (as is implied in the FAQ..."same looking effect"), then they should not be directly compared.

What bothers me, is that MadOnion goes out of their way in the FAQ to explain that they built the test for "compatibility."

FAQ:
The Advance Pixel Shader test is what we call a Feature Test, which means that we, above all, want to present some new technology.
Click to expand...

There is a conflict between showcasing "new technology" and making it "compatible" with old technology. Which is it? A showcase of the new PixelShader 1.4 technology, or not?

The point of "showcasing technology" is to do something specifically WITHOUT regard to compatibility with older tech. When you do that, you can be limiting what you're doing with the new tech.

Why not have one bump-mapping feature test. Fall-back to emboss mapping if Dot-3 isn't present. You can generate two "scenes" that show Emboss looking pretty close to Dot-3 for "compatibility", but only if you limit how you present DOT-3.

FAQ:
Both modes could be optimized to show more performance either way, but now the test is just optimized for maximum compatibility.
Click to expand...

So with this in mind, isn't it actually DANGEROUS to show scores for this "feature" test? I mean, it sounds like very little consideration for performance was given, and most was on making it "compatible".

It is unbelievable to me, that at seemingly EVERY opportunity MadOnion has to make a decision, where deciding one way would "favor" nVidia and the other way would "favor" someone else....the decision has ALWAYS been in favor of nVidia.

I'm not one to typically promote conspiracy theories and hidden alliances, but hell, Mad Onion makes it almost impossible to believe that nVidia doesn't have extremely heavy influence on how 3D Mark is presented.
 
Why am I not surprised? :rollseyes:

Oh, what makes the "advanced pixel shader" test advanced? It supposedly can be done with PS1.0 ...
 
Come on people. MadOnion includes a specific test for PS1.4 and they are "biased" against ATI because they also render the same thing using PS1.0? Geez. I can't even running the EMBM test on a GF2.


Are they biased in favor of Kyro because 3DMark 2001's rolling demo provides emulation of the nature scene using DX7 hardware?

The fact is, there are very few effects that can be done with PS1.4 that can't be done in PS1.0 and most of them are related to dependent lookups. Madonion would be hard pressed to code an effect that can't done in a different way on other cards (dependant lookups is one of them)

Perhaps they should be chided for artistic unoriginality, but bias? Give me a break.
 
Who said anything about effects Democoder, 3Dmark is a PEFORMANCE ANALYZER, even in the bottom corner "frames is everthing". OK MADONION talk the talk, Pixel Shader 1.4 when implemented properly will be FASTER. 3Dmark is a benchmarking utilty, we want the fastest frames possible...so in other words do it right or don't DO IT AT ALL. If I want to see what Pixel Shader 1.4 can do all I need is the technology demos from ATI, the Nature test with transparent water blurring the swimming fish below...
LOL since when did 3Dmark become a demo for new technology..garbage.
 
Perhaps the wording in the documents could have been clearer maybe but the idea behind the APS test is to show the performance difference that PS1.4 has over earlier revisions by having the 2 phases per pass - this is what makes it "advanced".

The test uses the same routines in PS1.0 as it does in PS1.4 (well, the same visual effects if you know what I mean) but the latter can take advantage of the phase instructions.

The claims of bias are so tired now it beggars belief at times. Hey aren't all games biased too? :rollseyes: After all nearly every single one of them only run on one operating system; can't use them for benchmarks in reviews then.
 
You must log in or register to reply here.

Similar threads

R
3DMark - Solar Bay Ray Tracing test available now
Replies
7
Views
2K
Lurkmass
L
orangpelupa
How to compare detailed 3D Mark benchmark result?
Replies
0
Views
682
orangpelupa
orangpelupa
D
Video Game Console Cooling - A Hot Topic (pun kind of intended).
Replies
14
Views
3K
Shifty Geezer
Shifty Geezer
Rys
Reopening Architecture and Products + rules and guidelines changes
3 4 5
Replies
90
Views
16K
Shifty Geezer
Shifty Geezer
N
Deja Vu from the past :)
2
Replies
29
Views
5K
function
F
Back
Top