Post your 3dmark05 results

[gururu]

Originally posted by: HaloEighty8

Originally posted by: lordtyranus

Originally posted by: HaloEighty8
5824 3DMarks

Damn. Really nice for stock speeds.



Thanks, I honestly didn't think I would get a score that decent for stock hardware.

i didn't know you could run an x800xt in a sff box. how much power does your power supply have? i 've been daydreaming of such a box.

 

[blckgrffn]

This says it all about 05:

http://service.futuremark.com/compare?3dm05=19126

Pretty damn funny, I think....

I am still waiting for the huge file to download, can't find a decent mirror...

 

[gururu]

Originally posted by: blckgrffn
This says it all about 05:

http://service.futuremark.com/compare?3dm05=19126

Pretty damn funny, I think....

I am still waiting for the huge file to download, can't find a decent mirror...

i think his cpu mark is more ridiculous than his 3dmark!!!

 

[HaloEighty8]

Originally posted by: gururu

Originally posted by: HaloEighty8

Originally posted by: lordtyranus

Originally posted by: HaloEighty8
5824 3DMarks

Damn. Really nice for stock speeds.



Thanks, I honestly didn't think I would get a score that decent for stock hardware.

i didn't know you could run an x800xt in a sff box. how much power does your power supply have? i 've been daydreaming of such a box.

Yeah, it actually works fine.. it's kind of cool to have all the power in this little box. It's a Shuttle SB75G2 i875 based with a 250w p/s.. just in case you where wondering. I'm running dual 74g raptors in it as well.. with the shuttles, power is never the problem.. it's heat!

 

[n7]

4379

That's with my Leadtek 6800 GT OCed to Ultra speeds using the offical Nvidia drivers...
It seems the unofficial drivers users are getting higher scores.

 

[user1234]

Originally posted by: PrayForDeath

Originally posted by: apoppin

Originally posted by: lordtyranus

Where? not YOUR link

Well, someone scored 7k 3dmarks with a single XTPE. Very close to the SLI setup.

Perhaps it is "made-up" . . .. techreport is only getting 5771
:roll:

Acutally, the users in Futuremark's website were able to achieve 7726 on an XT PE, but guess what, they overlcocked the hell out of it! 726 Core, 641 Memory!!!!!!
There are also people who got 7695, 7297 and 7225 all on overlcocked XT PEs, which means their overclocked XT PEs are faster than SLI 6800Ultras! I think SLI Ultras will score higher than the Inquirer claimed.

ummm..... I guess you don't get it...scoring higher on 3dcrap05 doesn't mean it's faster in the real world, all it means is that it's faster on whatever stupidmark is testing, which is some harldy relevant crap.

 

[Bateluer]

I scored a 3700 on a P4C 2.4Ghz with a 6800 GT at stock speeds. Is it just me or does the Free version skip all the shader tests? It only runs the Game and CPU benches, then spits out a score.

 

[user1234]

Originally posted by: CaiNaM
i'm not sure why everyone disses 3dm so much.. it certainly is not the "difinitive" scale of performance (no more so than any other single game). 3dm2003 results have shown it's results are pretty consistent with d3d games such as far cry, so it's certainly no less relevant than any other single app.

The highlighted claim is pure bs, 3dcrap03 (kust like05) only measures GPU performance, so it's only relevant in comparing different graphics subsystems used on the same machine, which is what hardware round-up comparisons usually do. But it's completely irrelavant in predicting expected frame rates in real world games on machine A relative to machine B. Unfortunately for you and Stupid Mark, real world performance depends on a whole lot more factors than just GPU subsystem.

 

[gururu]

Originally posted by: user1234

Originally posted by: CaiNaM
i'm not sure why everyone disses 3dm so much.. it certainly is not the "difinitive" scale of performance (no more so than any other single game). 3dm2003 results have shown it's results are pretty consistent with d3d games such as far cry, so it's certainly no less relevant than any other single app.

The highlighted claim is pure bs, 3dcrap03 (kust like05) only measures GPU performance, so it's only relevant in comparing different graphics subsystems used on the same machine, which is what hardware round-up comparisons usually do. But it's completely irrelavant in predicting expected frame rates in real world games on machine A relative to machine B. Unfortunately for you and Stupid Mark, real world performance depends on a whole lot more factors than just GPU subsystem.

ok someone had to say it... your spewing nonsense. firstly, your theory doesn't include why many people with different cpu and system configurations are pretty much getting the same scores as what the reviews are getting (with a given gpu at stock settings). In 3dmark05 at least, current system configurations hardly impact the score much at all. secondly, Cainam's generalization hasn't been shown to be incorrect. d3d games sort out gpus about as well as 3dmark03 did in every case I'm familiar with. Different system configurations certainly impact game performance, but the gpu pecking order generally remains the same. And that's all 3dmark tries to establish.

 

[exdeath]

I only get 5435

FX53 @ 2.4 Ghz
MSI K8N Neo2 Platinum
1024 MB Corsair XL 3200 @ 2-2-2-5-1T
eVGA 6800 Ultra 425/1100

But I get 85.4 fps in Doom 3 @ 1600 x 1200 Ultra quality demo1

Funny how I always get lower scores in stuff like 3DMark, but I can run real games faster than people getting 1000 3d marks more than my machine hah.

 

[gururu]

Originally posted by: exdeath
I only get 5435

FX53 @ 2.4 Ghz
MSI K8N Neo2 Platinum
1024 MB Corsair XL 3200 @ 2-2-2-5-1T
eVGA 6800 Ultra 425/1100

But I get 85.4 fps in Doom 3 @ 1600 x 1200 Ultra quality demo1

Funny how I always get lower scores in stuff like 3DMark, but I can run real games faster than people getting 1000 3d marks more than my machine hah.

well, your 3dmark beats my score, so it predicted that you'd beat me in Doom 3 too!

 

[yezhou]

Originally posted by: exdeath
I only get 5435

FX53 @ 2.4 Ghz
MSI K8N Neo2 Platinum
1024 MB Corsair XL 3200 @ 2-2-2-5-1T
eVGA 6800 Ultra 425/1100

But I get 85.4 fps in Doom 3 @ 1600 x 1200 Ultra quality demo1

Funny how I always get lower scores in stuff like 3DMark, but I can run real games faster than people getting 1000 3d marks more than my machine hah.

Based on thesebenchmarks, your score of 5435 looks pretty good! I only get 37xx-39xx and I basically have the same machine as yours!
BTW. what IQ settings in the NVidia settings are you using with Doom 3? I only get 82.7fps at 1280x1024 High Quality w/ noAA and 8AF with "Quality" IQ under NVidia settings.

 

[CaiNaM]

Originally posted by: user1234

Originally posted by: CaiNaM
i'm not sure why everyone disses 3dm so much.. it certainly is not the "difinitive" scale of performance (no more so than any other single game). 3dm2003 results have shown it's results are pretty consistent with d3d games such as far cry, so it's certainly no less relevant than any other single app.

The highlighted claim is pure bs, 3dcrap03 (kust like05) only measures GPU performance, so it's only relevant in comparing different graphics subsystems used on the same machine, which is what hardware round-up comparisons usually do. But it's completely irrelavant in predicting expected frame rates in real world games on machine A relative to machine B. Unfortunately for you and Stupid Mark, real world performance depends on a whole lot more factors than just GPU subsystem.

heh.. what i'm saying is nonsense? umm.. take a look at what you posted.

of course performance on machine a relative to machine b will be different. and performance in anything depends on other things as well as the gfx subsystem. this concept may be nonsensical to you, but bear with me a minute: the whole point of a benchmarking gfx cards is to keep EVERYTHING THE SAME EXCEPT THE GFX SUBSYTEM, thereby attempting to measure the differences of just the card. if you don't have a standard "baseline", it has no relevance whether you're using a "game" or a synthetic benchmark. doh!

when saying performance in 3dm reflects perf in many d3d games (and they often do if you took the time to look), what is reflective is the performance of the cards relative to each other, not 3dm vs this game or that game...

 

[user1234]

Originally posted by: gururu

Originally posted by: exdeath
I only get 5435

FX53 @ 2.4 Ghz
MSI K8N Neo2 Platinum
1024 MB Corsair XL 3200 @ 2-2-2-5-1T
eVGA 6800 Ultra 425/1100

But I get 85.4 fps in Doom 3 @ 1600 x 1200 Ultra quality demo1

Funny how I always get lower scores in stuff like 3DMark, but I can run real games faster than people getting 1000 3d marks more than my machine hah.

well, your 3dmark beats my score, so it predicted that you'd beat me in Doom 3 too!

Look, you said it yourself, 3dmark only seems to test the graphics subsystem alone - it's a purely GPU bound test score. But most games are CPU bound, especially when using high end video cards. For example, if you look at this article about doom 3 performance, you shall see that Athlon XP 2500+ (@1.83 GHz) has about half the frame rate as Athlon 64 FX-53, when both use the 6800 Ultra. This is a real world performance benchmark. Now let's look at 3dmark - exdeath system (above) scores 5435, while an athlon XP 2500+ with the same 6800U GPU scores about 5300, which is only about 3% less. And in fact if you'd overclock the GPU used with the Athlon XP just a little its score will surpass the FX-53's score, but we know very well there isn't any game in existence in which the Athlon XP system will produce even two thirds the frame rate of the FX-53 system. Therefore, 3djunk is not a good predictor of a system's real world performance in gaming. In other words, just because you score lower (or higher) than another system doesn't mean your gaming performance (measured by cold hard FPS rate) is worse (or better). In fact, the correlation is very weak, because (again) 3dcrack only measures the graphics subsystem. I've studied the perfromance of many systems and usually a system'ss performance can be reliably predicted by looking at its detailed specs and the characteristics of each component. But in many many cases, I've seen 3djack scores which are just completely out of whack with what kind of gaming performance you'd expect to get from a system based on it's overall specs. Ergo, there is very little value in comparing scores between different systems across the internet, like you've been doing in this thread.


[pwned]

 

[oldman420]

2655 stock

 

[gururu]

Originally posted by: user1234

Originally posted by: gururu

Originally posted by: exdeath
I only get 5435

FX53 @ 2.4 Ghz
MSI K8N Neo2 Platinum
1024 MB Corsair XL 3200 @ 2-2-2-5-1T
eVGA 6800 Ultra 425/1100

But I get 85.4 fps in Doom 3 @ 1600 x 1200 Ultra quality demo1

Funny how I always get lower scores in stuff like 3DMark, but I can run real games faster than people getting 1000 3d marks more than my machine hah.

well, your 3dmark beats my score, so it predicted that you'd beat me in Doom 3 too!

Look, you said it yourself, 3dmark only seems to test the graphics subsystem alone - it's a purely GPU bound test score. But most games are CPU bound, especially when using high end video cards. For example, if you look at this article about doom 3 performance, you shall see that Athlon XP 2500+ (@1.83 GHz) has about half the frame rate as Athlon 64 FX-53, when both use the 6800 Ultra. This is a real world performance benchmark.Now let's look at 3dmark - exdeath system (above) scores 5435, while an athlon XP 2500+ with the same 6800U GPU scores about 5300, which is only about 3% less. And in fact if you'd overclock the GPU used with the Athlon XP just a little its score will surpass the FX-53's score, but we know very well there isn't any game in existence in which the Athlon XP system will produce even two thirds the frame rate of the FX-53 system. Therefore, 3djunk is not a good predictor of a system's real world performance in gaming. In other words, just because you score lower (or higher) than another system doesn't mean your gaming performance (measured by cold hard FPS rate) is worse (or better). In fact, the correlation is very weak, because (again) 3dcrack only measures the graphics subsystem. I've studied the perfromance of many systems and usually a system'ss performance can be reliably predicted by looking at its detailed specs and the characteristics of each component. But in many many cases, I've seen 3djack scores which are just completely out of whack with what kind of gaming performance you'd expect to get from a system based on it's overall specs. Ergo, there is very little value in comparing scores between different systems across the internet, like you've been doing in this thread.

[pwned]

After reading your thoughts, I don't think we really disagree. You are stating a point for what 3dmark does not accomplish (doesn't show how cpu/memory dependent a game is) and I am stating a point about what 3dmark tries to accomplish (informs one of which gpu is fastest in D3D). The problem I think you have is though, is that there is no benchmark or timedemo on the planet that can definitively predict what another application requires. And I don't think many people will believe a prediction someone might make based on an opinion regardless of the experience they may have in reviewing system configurations.

 

[DAPUNISHER]

CaiNaM,

There's good logic in what your stating, I'm just disenchanted and "acting out" Having hit a hotdeal of $299.99 for a BFG GT without the assistance of 3DM in the evaluation and decision process to purchase, leaves me convinced that I do not need their bench for any reason whatsoever other than to see the impressive eye candy. Yes, it's a preview of things to come when it first shows up, but for me that is where it's value ends. Particularly since by the time a number of games *especially good ones* that load the card as heavy as '05 does, hit the shelves, I'll likely need a new card anyways. Thus I've concluded most of it's benefits are dubious at best, at least in my case I will concede it has its uses, but I still feel it's not deserving of the significance many attach to it.

lordtyranus,

It is odd looking@our scores and system specs side by side that my score isn't on par with your's. My FPS in games are always within a few percent of the review sites using a setup very similar to mine but with faster timings so I'm not certain what the issue is with '05 I'll try a few tweaks that aren't time intensive and see what happens for sh!ts&giggles

 

[DAPUNISHER]

Originally posted by: JBT

Originally posted by: CaiNaM

Originally posted by: DAPUNISHER
Oh yeah, forgot to post mine 4904

spanks mine

you must beat me in the cpu scores due to that 1 mhz clockspeed advantage! you have nc or ch cpu?

From my initial testing of this the core speed really helps scores. My first run @ 400/1100 gets a score of 43XX with the older 65.xx drivers, updating to the 66.70's and OCing to 423/1140 really improves the score a lot 5375

Yep, I picked up 900 points just going from the 61.77 to 66.70. I have a 3000+ that is nearly 10 months old, the board is 9, and the ram is probably close to 2yrs old now.

 

[HumblePie]

user1234.... grrr... are you that dumb? Seriously? I mean not to flame like that but I can only read so much ignorance and stupidity before I feel like I'm about to lose IQ points.

Look. 3d mark 2001se is all about CPU. Changing from a ti4200 to an x800 xt pe or 6800 ultra with everything else being the same in the system will not change the overall score that much. However, taking a cpu up to a higher clock rate, or the front side bus with the memory higher with the everything else in the system remaining the same and you'll see a HUGE difference in scores. This is a decent representation of system performance and no GRAPHICS performance. Many people who are wanting to test GRAPHIC performance were severly pissed that a benchmark for graphics would vary much more based on CPU/memory speeds then the actual graphic card used.

Now 3d mark 2003 was a little bit of a mixture. Changing either CPU, memory or graphics card will give a noticeable difference. There is some CPU bottlenecking going on though with the HIGHER END, as in the new R4xx+ and NV4x+ line of graphic cards that came out. With these cards, cpu and memory bottlenecking come back into play.

So now people want to see differences in higher end graphic cards. So 3d mark 2005 aims to do that. By stressing the crap out of the capabilities of these cards. Which is why anything below a x800 or 6800 is basically like watching a slide show. One can easily delinate how well using THIS standardized test for D3D will show performance between high end graphic cards with the rest of the system remaining close to equal in CPU/memory speeds.

Now does everyone play D3D games? not always. But the perponderous of games out there are D3D with OpenGL following close in behind.

These tests are meant to stress ONE THING and ONE THING ONLY from the test. If you want to see how well your system performs overall you have to test everything.

This is the same for all tests. You don't base your car buying decisions based on how fast a car can get from 0-60mph do you? Or its safety rating? or torque? No. Cars, like computers come in wide varieties and you have to realize WHAT YOU WANT TO DO with your piece of equipment. Then use the appropriate tests to rationalize your decision to go with one peice of hardware over another.

If you play Doom3 get an nvidia card and shut up. If you play D3D games or maybe the upcoming half life 2 game then buy ATI and shut up. If you want a card to play any game now, and many games coming ou the conceivable future then there are a ton of cards to select from unless you are looking for specific frame rates with certain eye candies turned on. Then you have to buy the high end cards from either ATI or Nvidia. In either case, if you want your system to play games the best you buy an AMD A64. If you like to do video encoding or decoding and want a PC you get a pentium. Otherwise get a Mac. I could go on and on and on.


Bottom line. This is a TOOL, and like all tools, it is only as good as you use it. You can't get pissed off at a hammer and call it crap because it can't cut a hole out of a sheet of metal. I say that because this is EXACTLY what you are doing when you say 3d mark 2005 sucks and is craptastic. It's a tool and does one thing and one thing only. And it does that one thing VERY WELL. Learn to realize that and stop spewing crap out your anus.

 

[n7]

Originally posted by: DAPUNISHER

Originally posted by: JBT

Originally posted by: CaiNaM

Originally posted by: DAPUNISHER
Oh yeah, forgot to post mine 4904

spanks mine

you must beat me in the cpu scores due to that 1 mhz clockspeed advantage! you have nc or ch cpu?

From my initial testing of this the core speed really helps scores. My first run @ 400/1100 gets a score of 43XX with the older 65.xx drivers, updating to the 66.70's and OCing to 423/1140 really improves the score a lot 5375

Yep, I picked up 900 points just going from the 61.77 to 66.70. I have a 3000+ that is nearly 10 months old, the board is 9, and the ram is probably close to 2yrs old now.

Holy sh!t!

I noticed how everyone was getting higher scores with the 66.70 drivers, so installed them...& i went from a score of 4379 to 5239!

860 points difference! I have a funny feeling that some quality is being sacrificed though to get such a huge jump.

 

[FluxCap]

Originally posted by: lordtyranus

Originally posted by: DAPUNISHER
Oh yeah, forgot to post mine 4904

Why so low?

Why so high you mean?

@Punisher,

How the heck is yours so high with a GT? What overclock do you have on that thing? I only got 3900!

Edit: Just saw your comment:

Yep, I picked up 900 points just going from the 61.77 to 66.70. I have a 3000+ that is nearly 10 months old, the board is 9, and the ram is probably close to 2yrs old now.

I had no idea drivers would make that kind of jump, something smells fishy!

 

[user1234]

HumblePie and gurururururu.... I think we all understand what this benchmark does and we don't disagree on what the score means. But the point I'm trying to convey is about the relevance and importance of such a test. We all agree that 3dm5 does not predict overall system perofrmance, as you said it's just a tool that measures one thing. When evaluating a system people want to know how it will perform, but obviously a general evaluation cannot measure performance in every possible game and app, because that's an almost infinite number. Therefore, most reviews use a number of measurements, and each of these measurements are chosen to be the representative of a class of games and/or apps, and to give a sense of the real world performance of the machine for that class of games/apps. That is, they try to choose the most useful benchmarks. Now, let's see what's the usefulness of 3d-shark5 - as we all agree it only measures the GPU, so given that you know the machine video card's spec, you will already pretty much know what it would score in 3dmark05. It's kind of like measuring the hard disk transfer rate - it's usually the same if you use the same hard disk. Therefore, most professional reviewers (see: Maximum PC, PC Gamer) have decided not to use this benchmark when evaluating a PC system (not when comparing graphics cards), since again it's just basically a technical measurement of one component, not the whole system, and it ends up being more confusing because it can give opposite results to real world tests. So bottom line, there's very little reason to use this test to evaluate system gaming performance, but many people (example: caiman) still think this score represents the FPS rate they will get in games like far cry.

[pwned]

 

[EviLRas]

ouch! I have Amd 64 3200+, 1gb pc3200 Geil Ram, Asus K8vSE MoBo, and Ati x800 PRO. everything running at stock, as dont trust overclocking! just using the overdrive feature on the ati application. scores as follows:


20546.00 3d mark
9940.00 3dmark04
3647 3dmark05
52345 aquamark3.0


3dmark05:
GT1 - 15.4fps
GT2 - 10.0 fps
GT3 - 20.2fps

CPU1 - 1.8fps
CPU2 - 3.0fps

I'm not convinced that 3647 is a respectable score, considering my PC cost me £1200 new!!!
I'm feeling Very naked with this system now...


Rascle


:shocked:

 

[EviLRas]

ps. dont really mind about what the scores say, i'm happy with the way my system handles games. so i figure thats more important!


Rascle

 

[Marsumane]

Originally posted by: user1234
HumblePie and gurururururu.... I think we all understand what this benchmark does and we don't disagree on what the score means. But the point I'm trying to convey is about the relevance and importance of such a test. We all agree that 3dm5 does not predict overall system perofrmance, as you said it's just a tool that measures one thing. When evaluating a system people want to know how it will perform, but obviously a general evaluation cannot measure performance in every possible game and app, because that's an almost infinite number. Therefore, most reviews use a number of measurements, and each of these measurements are chosen to be the representative of a class of games and/or apps, and to give a sense of the real world performance of the machine for that class of games/apps. That is, they try to choose the most useful benchmarks. Now, let's see what's the usefulness of 3d-shark5 - as we all agree it only measures the GPU, so given that you know the machine video card's spec, you will already pretty much know what it would score in 3dmark05. It's kind of like measuring the hard disk transfer rate - it's usually the same if you use the same hard disk. Therefore, most professional reviewers (see: Maximum PC, PC Gamer) have decided not to use this benchmark when evaluating a PC system (not when comparing graphics cards), since again it's just basically a technical measurement of one component, not the whole system, and it ends up being more confusing because it can give opposite results to real world tests. So bottom line, there's very little reason to use this test to evaluate system gaming performance, but many people (example: caiman) still think this score represents the FPS rate they will get in games like far cry.

[pwned]

Im going to stay out of this rediculous argument. But basically I have to say this:
Using "pwned" at the end of every rant usually results in an unconstructive argument due to the reactions of the people that it is aimed towards. You really should post your points on more of a discussion level as opposed to an argumentative offensive.