3Dmark 05 = blah.

[TGHI]

Yet another rant about 3Dmark05, but when I ran this 'benchmark' for the first time yesterday, I was just plain insulted. How can a systen that gets 45000 Aquamarks get 1100 in 3dmark05? Sure, the tests were beautiful, but they ALL ran at less than 2FPS the ENTIRE time. What is Futuremark up to? Do they think everyone has an Athlon64 with a 6800 Ultra? Will even THAT setup run smoothly?

Looks to me like they are trying to get people to buy a professional license so that they can run it at a lesser resolution, considering that I'm willing to bet that < %90 of all computers can't handle the benchmark smoothly.

I know that I only have a lowly 5900xt, which is considered a mid-range card - but I felt like I was running a TNT2 M64....what gives? It should be renamed to 'the fat-wallet' benchmark. Pox on you, Futuremark.

 

[EightySix Four]

It's designed to run like that, they want it to bring even the highest of the high end card to their knees, because that's what future games will do...

 

[THUGSROOK]

Originally posted by: TGHI
Looks to me like they are trying to get people to buy a professional license so that they can run it at a lesser resolution, considering that I'm willing to bet that < %90 of all computers can't handle the benchmark smoothly.

let me assure you ~ running 3dm05 at a lower res DOES NOT HELP.

start\settings\control panel\add remove programs\3dmark05\uninstall...

 

[blurredvision]

You're missing the point of the 3DMark program. By releasing such a stressful benchmark, it will test even the newest, fastest systems for the next 2-4 years. IMO, 3DMark01 is now outdated, as the scores have gotten outrageous, but 3DMark03 is still viable for atleast another round of cards.

You say you felt like you were running a TNT2? Well, get used to it for the games coming in the next couple of years, because that's what this benchmark attempts to emulate. You said it yourself, you have a lowly 5900xt. Nothing more, nothing less.

You look at this benchmark like it should give you some holy grail of a number for your card, when the program is merely a benchmark to test your card for the upcoming games. If you got a low score, then don't plan on playing the AAA, graphic-intense games coming in the next year or two. Simple as that.

 

[n7]

My A64 + 6800 GT looked like crap with 3Dmark05...but who cares; it's just a benchmark.

 

[FREEZrBURN]

Look, the technology showcased in 3dmark05, is the technology that will be implemented in the near future. Obviously, The most top of the line system, will not run games released next year at the optimum settings. Hell, most top-end machines struggle with doom3's effects cranked up. Games, are also optimized for a stable fps. If you recall the opening test, you saw doom3 quality graphics shoved into a room with 500 things going on at once. Something that would never occur in game because the hardware cannot handle it. 3dmark05 is precedenting future games, not just current ones.

 

[bigwayne]

3DMark05 is like the game you bought when it was new and shelved for 2 years because it won't run on your top of the line hardware.

Take a look back after one-two years of hardware upgrades, and I'll bank you'll be pulling '03 scores.

As it is right now, no current hardware setup is optimized at all to run '05 (thats the idea), so its not even viable yet as a benchmark.

 

[CraigRT]

I agree with the others, it's suppose to be like that... no worries

 

[effee]

run 3dmark03, it'll make you feel better

 

[kamranziadar]

WELL I GOT 3010, i am not disappoint

 

[Pete]

Yet another rant about 3Dmark05, but when I ran this 'benchmark' for the first time yesterday, I was just plain insulted. How can a systen that gets 45000 Aquamarks get 1100 in 3dmark05?
Doesn't AM3 just multiply your average framerate by 1,000 to arrive at an official score? If it makes you feel better, just multiply your 3DM05 score by 10.

IIRC, AM3 is mainly SM1.x, with a sprinkle of SM2.x. 3DM05 is all SM2.0+, so expect it to be much harder on your hardware simply because it's doing so much more (the whole point of higher Shader Models is to allow for more flexibility in the calculations that create the pictures you see, and obviously FM is taking advantage of that extra headroom).

 

[Zebo]

start\settings\control panel\add remove programs\3dmark05\uninstall...

 

[Concillian]

GeForceFX series + Pixel Shader 2.0 = crappy performance

3dMark05 uses Pixel Shader 2.0 pretty heavily

therefore

GeForceFX series + 3dMark05 = crappy performance

I'd tend to agree that 3dMark programs aren't worth a whole lot, but in this case they ARE pointing out a major deficiency with the GeForceFX series that is being confirmed by newer games that use PS 2.0.

 

[RadiclDreamer]

I only got a 4364 and I have a 6800 Ultra OC and a P4 3000mhz precott and 2 GB ram. Im not complaining

 

[Acanthus]

3dmark is trash since 2001.

The 2001 version used real game engines. They completely abandoned that with the 2003 benchmark, and repeat their mistake in 2005. In no way does 3dm'05 represent ANY future game being released. As 2003 didnt.

 

[AnonymouseUser]

Originally posted by: Concillian
GeForceFX series + 3dMark05 = crappy performance

Actually, it's GeForceFX series + DirectX 9 = crappy performance

Yet GeForceFX series + OpenGL = respectable performance

Amazing what Microsoft can do to level the playing field, isn't it?

 

[dguy6789]

microsoft has NOTHING to do with it. ATI just happens to spend more time making their cards run dx good then nvidia does. It is lazyness on nvidia's part plain and simple.

 

[DrMindbender]

I don't really see it as a "gaming benchmark" as much as I do as a hardware benchmark or maybe a prosumer graphics benchmark. I mean there's so many different environmental factors involved you can't say THIS is what games are going to be published by game companies. I also think it's funny that Futuremark gets so much crap and review sites say "well we have our doubts about this test, but everyone else is using it blah, blah, blah, we're going to make it a key part of our benchmarking anyway. The only people that have to care about this test are NVIDIA/ATI and the wholesalers when they get together in the sales office.

 

[DrMindbender]

Well, didn't ATI stop trying to compete with NVIDIA back when hardware(openGL, glide, etc.) rendering was the coolest thing around and focus on software rendering(Direct3D) which at the time was considered to be the lame default? Then Microsoft actually improved DirectX and that changed everything. ATI became more successful because it was cheaper and easier to get games out in Direct3D because it was part of DirectX rather than designing games based off of all the different cards out there. That's not to say that ATI doesn't have powerful hardware, but the design is optimized for Microsoft as opposed to the game being optimized for the graphics card.

 

[OMG1Penguin]

3dmark is designed to scale, not to look good. By the time 3dmark07 comes out (which will not even be based on game-looking code, more than likely), it will run like trashxxors too.

 

[Regs]

Originally posted by: crazySOB297
It's designed to run like that, they want it to bring even the highest of the high end card to their knees, because that's what future games will do...

well in that case Ill wait until the future to use it.

 

[jim1976]

Originally posted by: Acanthus
3dmark is trash since 2001.

The 2001 version used real game engines. They completely abandoned that with the 2003 benchmark, and repeat their mistake in 2005. In no way does 3dm'05 represent ANY future game being released. As 2003 didnt.

What the man says... Don't even bother to see it as a benchmark for real gaming performance, just as a nice demo to watch and a point of reference for your system optimization.