Just how wacked is 3DMark03?

Well, if you look at how Futuremark characterizes their benchmark....

"By combining full DirectX®9.0a support with completely new tests and
graphics, 3DMark03 Pro continues the legacy of being the industry
standard benchmark. The high quality game tests, image quality tests,
sound tests and others give you an extremely accurate overview of your
system’s current gaming performance. "

you would think that it measures the performance of your entire
system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
http://www.aceshardware.com/read.jsp?id=60000242
for the lowdown.

"Folk" <> wrote in message
news:...
>
> Well, if you look at how Futuremark characterizes their benchmark....
>
> "By combining full DirectX®9.0a support with completely new tests and
> graphics, 3DMark03 Pro continues the legacy of being the industry
> standard benchmark. The high quality game tests, image quality tests,
> sound tests and others give you an extremely accurate overview of your
> system's current gaming performance. "
>
> you would think that it measures the performance of your entire
> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> http://www.aceshardware.com/read.jsp?id=60000242
> for the lowdown.

Yup. Last 3DMark that was any good was the 2001 edition, and even then it
has only really been good at comparing performance on the same system (like
when you change some BIOS settings etc).

/M

">>>>>" <> wrote in message
news:LcYVa.37640$ ...
> Because the f%#@$@%$ing vieocard used in each system is different!

Exactly. The cards HAVE to be the same for a test like that.


> Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!!
Helloo???
> "Folk" <> wrote in message
> news:...
> >
> > Well, if you look at how Futuremark characterizes their benchmark....
> >
> > "By combining full DirectX®9.0a support with completely new tests and
> > graphics, 3DMark03 Pro continues the legacy of being the industry
> > standard benchmark. The high quality game tests, image quality tests,
> > sound tests and others give you an extremely accurate overview of your
> > system's current gaming performance. "
> >
> > you would think that it measures the performance of your entire
> > system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> > out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> > http://www.aceshardware.com/read.jsp?id=60000242
> > for the lowdown.
> >
> >
>
>
>

On Wed, 30 Jul 2003 19:20:45 -0400, ">>>>>" <> wrote:

>Because the f%#@$@%$ing vieocard used in each system is different!
>Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!! Helloo???

You're missing the point. If 3DMark03 is supposed to be a test of the
*entire* system, then the 8 times more powerful P4 should have
compensated for the difference between two current technology video
cards.

Reading comprehension isn't your strong point, is it?


>"Folk" <> wrote in message
>news:.. .
>>
>> Well, if you look at how Futuremark characterizes their benchmark....
>>
>> "By combining full DirectX®9.0a support with completely new tests and
>> graphics, 3DMark03 Pro continues the legacy of being the industry
>> standard benchmark. The high quality game tests, image quality tests,
>> sound tests and others give you an extremely accurate overview of your
>> system's current gaming performance. "
>>
>> you would think that it measures the performance of your entire
>> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
>> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
>> http://www.aceshardware.com/read.jsp?id=60000242
>> for the lowdown.
>>
>>
>
>

"Folk" <> wrote in message
news:...
> On Wed, 30 Jul 2003 19:20:45 -0400, ">>>>>" <> wrote:
>
> >Because the f%#@$@%$ing vieocard used in each system is different!
> >Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!!
Helloo???
>
> You're missing the point. If 3DMark03 is supposed to be a test of the
> *entire* system, then the 8 times more powerful P4 should have
> compensated for the difference between two current technology video
> cards.
>
> Reading comprehension isn't your strong point, is it?

And 'how to do an experiment' isn't your strong point, is it?! For a proper
study you control variables so you are only testing one thing. If that site
had tested the 9600 and 9700 with the same CPU, there results would mean
something; as they are they don't.

No the P4 would not compensate for the difference. An 'entire' system
includes both video and CPU performance, despite what you think is going
with 3DMark03.


Too_Much_Coffee ®
>
>
> >"Folk" <> wrote in message
> >news:.. .
> >>
> >> Well, if you look at how Futuremark characterizes their benchmark....
> >>
> >> "By combining full DirectX®9.0a support with completely new tests and
> >> graphics, 3DMark03 Pro continues the legacy of being the industry
> >> standard benchmark. The high quality game tests, image quality tests,
> >> sound tests and others give you an extremely accurate overview of your
> >> system's current gaming performance. "
> >>
> >> you would think that it measures the performance of your entire
> >> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> >> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> >> http://www.aceshardware.com/read.jsp?id=60000242
> >> for the lowdown.
> >>
> >>
> >
> >
>
>

It's more of a test of the entire graphics system. If you put a crappy
video card in and expect a much faster/more-powerful to compensate then you
are going to be VERY disappointed.

Ron

"Folk" <> wrote in message
news:...
> On Wed, 30 Jul 2003 19:20:45 -0400, ">>>>>" <> wrote:
>
> >Because the f%#@$@%$ing vieocard used in each system is different!
> >Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!!
Helloo???
>
> You're missing the point. If 3DMark03 is supposed to be a test of the
> *entire* system, then the 8 times more powerful P4 should have
> compensated for the difference between two current technology video
> cards.
>
> Reading comprehension isn't your strong point, is it?
>
>
> >"Folk" <> wrote in message
> >news:.. .
> >>
> >> Well, if you look at how Futuremark characterizes their benchmark....
> >>
> >> "By combining full DirectX®9.0a support with completely new tests and
> >> graphics, 3DMark03 Pro continues the legacy of being the industry
> >> standard benchmark. The high quality game tests, image quality tests,
> >> sound tests and others give you an extremely accurate overview of your
> >> system's current gaming performance. "
> >>
> >> you would think that it measures the performance of your entire
> >> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> >> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> >> http://www.aceshardware.com/read.jsp?id=60000242
> >> for the lowdown.
> >>
> >>
> >
> >
>
>

"Dave" <> wrote in message
news:hX1Wa.25241$o%2.15185@sccrnsc02...

> All it really means is that 3Dmarket is a video showcase weighted a
> third for some DX9 functions, not even a really conclusive all-around
gaming
> benchmark at that. If it's intended to be representative of future DX9
> performance, I think it falls somewhat short on its own terms.

Oh, how I do like the thrust of this paragraph. '3DMarket'...brilliant.

Folk wrote:

> On Wed, 30 Jul 2003 19:20:45 -0400, ">>>>>" <> wrote:
>
> >Because the f%#@$@%$ing vieocard used in each system is different!
> >Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!! Helloo???
>
> You're missing the point. If 3DMark03 is supposed to be a test of the
> *entire* system, then the 8 times more powerful P4 should have
> compensated for the difference between two current technology video
> cards.
>
> Reading comprehension isn't your strong point, is it?
>

Nor is it yours.

You apparently don't understand that these tests are designed to stress primarily
the GRAPHICS hardware because THAT is what determines how well a system will play
a given game, much more than the CPU.

Case in point - I had a Matrox card that would just literally crawl in some
games. When I borrowed an nVidia based card and put it in the same machine, the
games literally flew.

So you can't compare performance of an "xyy" MHz system against a "zzz" MHz
system *unless* they contain the same graphics card.

If you don't compare against the same graphics card, in a graphics benchmark,
then all bets are off.

--Keith




>
> >"Folk" <> wrote in message
> >news:.. .
> >>
> >> Well, if you look at how Futuremark characterizes their benchmark....
> >>
> >> "By combining full DirectX®9.0a support with completely new tests and
> >> graphics, 3DMark03 Pro continues the legacy of being the industry
> >> standard benchmark. The high quality game tests, image quality tests,
> >> sound tests and others give you an extremely accurate overview of your
> >> system's current gaming performance. "
> >>
> >> you would think that it measures the performance of your entire
> >> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> >> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> >> http://www.aceshardware.com/read.jsp?id=60000242
> >> for the lowdown.
> >>
> >>
> >
> >

On Wed, 30 Jul 2003 21:11:41 -0700, "Too_Much_Coffee ®"
<> wrote:

>
>"Folk" <> wrote in message
>news:.. .
>> On Wed, 30 Jul 2003 19:20:45 -0400, ">>>>>" <> wrote:
>>
>> >Because the f%#@$@%$ing vieocard used in each system is different!
>> >Radeon 9600 on the P4 system against 9700 PRO on the P2 system!!!
>Helloo???
>>
>> You're missing the point. If 3DMark03 is supposed to be a test of the
>> *entire* system, then the 8 times more powerful P4 should have
>> compensated for the difference between two current technology video
>> cards.
>>
>> Reading comprehension isn't your strong point, is it?
>
>And 'how to do an experiment' isn't your strong point, is it?! For a proper
>study you control variables so you are only testing one thing. If that site
>had tested the 9600 and 9700 with the same CPU, there results would mean
>something; as they are they don't.

They were not trying to compare the performance differences of two
video cards. Hardware enthusiast sites do that on a regular basis....
we've all seen hundreds of them. How do you prove that a benchmark is
*not* a system benchmark but a video card benchmark? You take two
video cards that are similar in performance and use them in two *very*
dissimilar hardware platforms. If 3DMark03 were truly a system
benchmark, then the eight times more powerful P4 would compensate for
a video card that is what.... 15% faster. But it didn't.

>No the P4 would not compensate for the difference. An 'entire' system
>includes both video and CPU performance, despite what you think is going
>with 3DMark03.

Why does that sound like two contradictory statements? Yes, an entire
system does include video, cpu, memory and hard drive performance. If
3DMark03 was a system test, then System A with a 100Mhz cpu and a
200Mhz video card (just using numbers here) would be just as fast as
System B with a 200Mhz cpu and a 100Mhz video card. Not only did that
not happen here, but the system specs were *hugely* exaggerated while
the video card specs were just mildly different, and the dinosaur
still won.

That's the point.


>>
>>
>> >"Folk" <> wrote in message
>> >news:.. .
>> >>
>> >> Well, if you look at how Futuremark characterizes their benchmark....
>> >>
>> >> "By combining full DirectX®9.0a support with completely new tests and
>> >> graphics, 3DMark03 Pro continues the legacy of being the industry
>> >> standard benchmark. The high quality game tests, image quality tests,
>> >> sound tests and others give you an extremely accurate overview of your
>> >> system's current gaming performance. "
>> >>
>> >> you would think that it measures the performance of your entire
>> >> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
>> >> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
>> >> http://www.aceshardware.com/read.jsp?id=60000242
>> >> for the lowdown.
>> >>
>> >>
>> >
>> >
>>
>>
>
>
>

Interesting test. It basicly confirms what I already knew... 3DMark 2003
is a pile of crap and performance bears no resemblence to real games. Even
3DMark 2001 is pushing it. OTOH, I think Jedi Knight and Quake III are much
better benchmarks because they are real games and in my experience give
results that you can apply to real gaming. In real world gaming, CPU still
matters, and often a fast CPU with an older graphics card will outperform a
slow system with a fast graphics card.

"Folk" <> wrote in message
news:...
>
> Well, if you look at how Futuremark characterizes their benchmark....
>
> "By combining full DirectX®9.0a support with completely new tests and
> graphics, 3DMark03 Pro continues the legacy of being the industry
> standard benchmark. The high quality game tests, image quality tests,
> sound tests and others give you an extremely accurate overview of your
> system's current gaming performance. "
>
> you would think that it measures the performance of your entire
> system. Then why does a 350Mhz PII with 100Mhz SDRAM beat the crap
> out of a 2.8Ghz P4 with dual channel PC2100 DDR? Check out
> http://www.aceshardware.com/read.jsp?id=60000242
> for the lowdown.
>
>