Futuremark: 3DMark06

geo said:
If they aren't selling that solution across the AIB's, then I'd find that an inappropriate thing to do. Right now it is Dell-only, isn't it?

I don't think it is even availeble from Dell yet.
 
http://img.gamespot.com/gamespot/sha..._screen001.jpg
http://img.gamespot.com/gamespot/sha..._screen003.jpg
http://img.gamespot.com/gamespot/sha..._screen002.jpg

WoW

This is very interesting
CPU performance testing via AI and physics workloads for both single core systems and multi-threaded, multi-core and multiple processor environments

Anyway do someone knows here I will be able to get a video from 3Dmark06 in a system that can play it nice (eg a quad 7800s), so I can enjoy something that beautifull in motion.
 
tEd said:
:LOL:

Like they wouldn't do it just because it seems inappropriate

Ohhh, I don't like to assume that anyone is going to do a naughty until they actually do it. That wouldn't be fair. :LOL: Some people weren't/aren't overly happy about NV claiming the performance lead by comparing two-gpu/sli to single board solutions, but at least that form of SLI was/is broadly available across their partners.
 
pc999 said:
This is very interesting
CPU performance testing via AI and physics workloads for both single core systems and multi-threaded, multi-core and multiple processor environments
I can see what they are trying to do, but dont you think that 3DMark should be about testing the graphics card and only that? That the results should come from a test that is 100% or close to gfx card limited? Will not future shader and HDR heavy games be very GPU limited anyway on todays systems?
 
MistaPi said:
I can see what they are trying to do, but dont you think that 3DMark should be about testing the graphics card and only that? That the results should come from a test that is 100% or close to gfx card limited? Will not future shader and HDR heavy games be very GPU limited anyway on todays systems?

The individual tests will serve that purpose and people will soon learn that a fast CPU is necessary to obtain a high 06 score.
 
MistaPi said:
I can see what they are trying to do, but dont you think that 3DMark should be about testing the graphics card and only that? That the results should come from a test that is 100% or close to gfx card limited? Will not future shader and HDR heavy games be very GPU limited anyway on todays systems?

Future games will be the same as current games: both GPU and CPU limited depending on the situation. PC games designers will make use of all available resources to the extent that they can.

While I would prefer that the 3DMark score only be about GPU testing...realize that it appears there is a separate CPU "test" that contributes to the 3DMark 06 score. In other words, the "GPU game tests" should most definitely be GPU limited, whereas the separate CPU test will the CPU limited.

So, the end result is that the 3DMark 06 "score" will not simply be a measurement of of the GPU. However, we can still discern GPU power by looking at the individual GPU game tests in isolation. It'll be interesting to see how the press / reviewers handle this....I'm not optimistic that they will "get" the fact that the overall 3DMark score is not going to be a measurement solely of GPU performance. We'll see though. ;)

Edit: Ok, John beat me to it!
 
It'll be interesting to see how driver-CPU load is reflected in these benchmarks.

One thing that's received scant attention so far as I can tell is the different scaling with CPU-capability of ATI and NVidia graphics/drivers.

This has started to become really obvious with dual-core CPUs, particularly now as a 6800U is getting some insanely high scores in 3DMk05 as far as I can tell.

Jawed
 
It'll also be interesting if Ageia ever gets their hardware out the door...as I assume 3DMark 06 CPU test would take advantage of the "PPU" if one exists...
 
MistaPi said:
but dont you think that 3DMark should be about testing the graphics card and only that? That the results should come from a test that is 100% or close to gfx card limited?

Actualy I prefer a test that can tell me what is the best PC to play games overall isnted of the best gfx card alone, I would be very interesting a unified test too if possible.

Anyway probably one can disable that to test gfx only, as HW reviewer it should be better separated tests, and it still based on that phylosophy right.
 
I would soooo love to spill more beans, but rest assured, we haven't forgot the users/media who want to benchmark GPU performance only. You'll see soon what I am refering to. ;)
 
Nick[FM] said:
I would soooo love to spill more beans, but rest assured, we haven't forgot the users/media who want to benchmark GPU performance only. You'll see soon what I am refering to. ;)

Please...oh please if there is an option to see two different scores...I hope that they each have big, bold, red letters with arrows, circles, and sticky notes that says "Hey dumb ass journalist...THIS score means THIS, and THAT score means THAT."
 
Joe DeFuria said:
Please...oh please if there is an option to see two different scores...I hope that they each have big, bold, red letters with arrows, circles, and sticky notes that says "Hey dumb ass journalist...THIS score means THIS, and THAT score means THAT."

Only two? :devilish:
 
Joe DeFuria said:
Please...oh please if there is an option to see two different scores...I hope that they each have big, bold, red letters with arrows, circles, and sticky notes that says "Hey dumb ass journalist...THIS score means THIS, and THAT score means THAT."
And the dumbass journalist would reply with "Well, what's next, 5 different scores, like for GPU, CPU, memory, physics and GPU+CPU+etc=system in an all-encompassing 3DMark package?"
 
how about some compare url's from 06

AMD 64 X2 2609 MHz - ATI RADEON X850 XT PE - 2368 3DMarks
http://service.futuremark.com/compare?3dm06=38

AMD 64 X2 2615 MHz - X1800 Series - 3452 3DMarks
http://service.futuremark.com/compare?3dm06=44

AMD 64 X2 2610 MHz - ATI RADEON X1300 - 1322 3DMarks
http://service.futuremark.com/compare?3dm06=46

AMD 64 2409 MHz - NVIDIA GeForce 6800 Ultra - 2442 3DMarks
http://service.futuremark.com/compare?3dm06=47

AMD 64 X2 2621 MHz - NVIDIA GeForce 7800 GTX - 5552 3DMarks
http://service.futuremark.com/compare?3dm06=50

AMD 64 3200+ stock - Nvidia 7800GT stock - 3236 3DMarks
http://service.futuremark.com/compare?3dm06=52


looks like 1280x1024 is finally the default res instead of 1024x768.....finally....

interesting

it's released at 14:00 GMT
 
gravy said:
looks like 1280x1024 is finally the default res instead of 1024x768.....finally....
What?!

I totally disagree with this decision, on account that 1024x768 is still the crucial rez where AA and AF come into play and that if FM wants a higher default, they should stick to 1280x960.
 
gravy said:
AMD 64 X2 2609 MHz - ATI RADEON X850 XT PE - 2368 3DMarks
http://service.futuremark.com/compare?3dm06=38

AMD 64 X2 2615 MHz - X1800 Series - 3452 3DMarks
http://service.futuremark.com/compare?3dm06=44

AMD 64 X2 2610 MHz - ATI RADEON X1300 - 1322 3DMarks
http://service.futuremark.com/compare?3dm06=46

AMD 64 2409 MHz - NVIDIA GeForce 6800 Ultra - 2442 3DMarks
http://service.futuremark.com/compare?3dm06=47

AMD 64 X2 2621 MHz - NVIDIA GeForce 7800 GTX - 5552 3DMarks
http://service.futuremark.com/compare?3dm06=50

AMD 64 3200+ stock - Nvidia 7800GT stock - 3236 3DMarks
http://service.futuremark.com/compare?3dm06=52


looks like 1280x1024 is finally the default res instead of 1024x768.....finally....

interesting

it's released at 14:00 GMT
Sucks for people with 17" crts...
their eyes will be sore...
So how come the 1800xl scores so poorly(just a bit faster than a 7800gt despite having much more cpu power), and why is there compare urls out before the product?
 
Last edited by a moderator:
You might consider the 7800GTX score is an actually an SLI configuration.
 
Back
Top