How I wish somebody would host it.......Did I ever tell you all you are the best bunch of overclockers ever. *Cough*send me the file
Printable View
How I wish somebody would host it.......Did I ever tell you all you are the best bunch of overclockers ever. *Cough*send me the file
We don't have it (by we, I mean those of us that aren't reviewers or ATi/nV employees).
:slap:
I also noticed the 7800gt score is with a 430/1000 speeds. That gives me hope for my 7800gt clocked at 500/1250 w/zalman cooler.
added my signature now too :) .
Such monstrous stupidity....so now Shader Model decides what score you get ? I'd rather they made it an all SM3.0 score...
How retarded can you get...
Perkam
wow, looks like this benchmark has no love for ATI, thats for sure...
Owners of Sm 2.0 cards would rather they got left out then face this stupidity...Quote:
Originally Posted by Kunaak
Perakm
Oh, is that how it works? So if your videocard supports SM3.0 you automatically get an advantage over the same videocard that only supports SM2.0.Quote:
Originally Posted by perkam
Yup...which gives NO indication whatsoever of next generation graphics performance.Quote:
Originally Posted by shimq1
It seems that fears have been confirmed and that this is just a filler between 3dmark05 and 3dmark07....:(
Perkam
Damn, my dreams of my X800GTO2 beating my friends 7800GTX are over....sigh....:)Quote:
Originally Posted by perkam
pfft that dream was over before it started :D
dual core CPUs .... 3d06 just doubles the CPU score !!! lol !!! :p:
Perkam
except for performance in ut2007.Quote:
Originally Posted by perkam
anyways whatever happened to the scores from intel systems :confused: .
Intel Dual Core
http://service.futuremark.com/compare?3dm06=65
Its so stupifying....because of lame attempt at cpu dependency...An X1800XT w/ Dual core proc would come equal to a GTX 512MB with a single core proc...
P4's get off ok...Both single core:
P4 @ 2.66ghz = 700 points in cpu score, A64 @ 2.55ghz = 100 points...so Intel users can rejoice ;) ... Intel dual core users...aka fugger...can get ready to take top spot ;)
:rofl: :eleph: :rofl: :eleph: :rofl: :eleph: :rofl: :eleph: :rofl: :eleph: :rofl: :eleph: :rofl: :eleph: :rofl: :eleph:
Pentium D @ scores almost 2000 points for CPU score !!! TWICE AS MUCH AS AN A64 at 2.5ghz !!!!!
:p:
Perkam
why so suprised at this?Quote:
Originally Posted by perkam
isnt the same case in previous 3d marks?
take 3d03 for example dx7 card can only run GT1
SM3.0 benching is necessary now
it is an indication of the performance of card in the future (well at least in the near future b4 DX10 is released)
SM3.0 and SM2.0b are both Directx 9.0c ;) Hence my surprise...
Also, care to explain how an 820 Pentium D can possibly increase performance by 100% over an A64 in 3dmark ? Cpu score is a component of 3dmark, and is therefore a part of it....this benchmark is screwy to say the least.
Perkam
Dual core vs single...
hmmmmQuote:
Originally Posted by perkam
i hereby attached the formula of caculating the 3d06 scores
lets see how many weight are these cpu scores ;)
Quote:
The 3DMark Score is calculated using the following formula:
SM2.0 Score = 120 x 0.5 x (SM2 GT1 fps + SM2 GT2 fps)
HDR/SM3.0 Score = 100 x 0.5 x (SM3 GT1 fps + SM3 GT2 fps)
CPU Score = 2500 x Sqrt( CPU1 fps x CPU2 fps)
We'll define for clarity:
GS for SM3.0 hardware = 0.5 x (SM2S + HDRSM3S)
GS for SM2.0 hardware = 0.75 x SM2S
3DMark Score = 2.5 x 1.0/(( 1.7/GS + 0.3/CPU Score )/2)
i.e. 2.5 x weighted harmonic mean of GS and CPU Score.
GT1 fps means the average frame rate measured in SM2.0 graphics test 1. CPU fps means the frame rate measured in the CPU test.
wtf is up with this benchmark? It's not anywhere NEAR real worldQuote:
Originally Posted by perkam
X1800XT + AMD X2 @ 2.6ghz = 7800GT + Intel Pentium D @ 3.6ghz
http://service.futuremark.com/compare?3dm06=65
http://service.futuremark.com/compare?3dm06=34
:shakes:
In 3dmark06, interesting for a full run X1800XT doesn't run shader particals SM3.0 test http://service.futuremark.com/compare?3dm06=34Quote:
Originally Posted by perkam
That's an interesting find eva....but it doesn't affect the score...
The Only Apples to Apples comparison that can be made is if we either look at ONLY the SM2.0 scores or ONLY the SM3.0 scores and compare across platforms...
I suspect they made the new tests and it turned out they killed most of todays cards so they added some additional stuff to the score that wasn't representative of anything close to 3d performance...
Perkam
methinks this benchie is just emulating how games released this year are going to perform :cool: .example:ut2007.
well it's multithreaded.Quote:
Originally Posted by perkam
An FX-60 getting 1939 versus a Pentium D 820 at 3.5GHz getting 1829 sounds reasonable to me.
There's a difference between multi-threaded and illogically influenced by dual cores....this is the latter...I've seen and used programs that are SMP aware...this doesnt come close to even mimicking that software.
Perkam
would be interesting to see what X1900XTX doesQuote:
Originally Posted by perkam
UT has always been amd biased. Bah it sucks how the x1300p out performs the 6600gt im getting that does 11k in 03 :-/ Oh well, I dont really card, all I play is ut and css. When 07 comes out Ill go 1900Quote:
Originally Posted by RAMMAN
Probably more accurate of late 07 Unreal 3 engine based games at high detail.
So is it released yet???
interesting
Quote:
Originally Posted by tyellbee
Apparently 12pm est :(
Hm.. 6800U's whipping X800XT's because of SM3 support
One of the screenshots looks very nice
i reckon thats going to change with ut2007.plus its sm3 and sm2 only so sm2.0b hardware is going to take a plunge performance wise just like what we are seeing in 3dmark06.Quote:
Originally Posted by sabrewolf732
hah, futuremark coders making code as efficient or representative of code done by epic? Hah :slap:Quote:
Originally Posted by afireinside
Why would it change? Because futuremark said so? I don't see how this 1 "benchmark" is making people say intel p4's are the new good gaming cpu's :rolleyes:Quote:
Originally Posted by RAMMAN
I don't know what the problem is. I mean it seems fair to me. In that comparison that Perkam just posted the Intel platform was OCed, but the AMD one was not, so it's only fair that the Intel one has a better score.
As for the CPU, I would expect dual core to get 2x the performance because graphics is very parallisable.
Also, from EVA's chart it looks like dual core gets you a 75% boost, which is perfectly reasonable.
no because intels involved.hence why intel cpus will mysteriously perform better in multithreaded apps starting with quake4,s latest patch :slap: .Quote:
Originally Posted by sabrewolf732
6800 GT + Intel Pentium 4 3005 MHz
3DMark Score 2178 3DMarks
SM 2.0 Score 960 Marks
SM 3.0 Score 893 Marks
CPU Score 651 Marks
6800 GS + AMD Athlon(tm) 64 1800 MHz
3DMark Score 1933 3DMarks
SM 2.0 Score 807 Marks
SM 3.0 Score 766 Marks
CPU Score 705 Marks
Upon further review it looks like amd is better in 06
Um, the quake patch is multithreaded, thus AMD X2's also got boosts :rolleyes: :stick:Quote:
Originally Posted by RAMMAN
yes but intels boosts were bigger than amd,s.something to do with sse instructions...anyway enough of this ot discussion.
Link?Quote:
Originally Posted by RAMMAN
Also, is it just more or does 06 not seem that graphically impressive? Futuremark can suck a nut.
One thing I think is ridiculous, the best CPU available can't even get 1 FPS in the CPU test. I mean what is this supposed to be indicating? Future performance for 2006? Try future performance for 10 years from now. This isn't telling you anything about current hardware performance. It would be one thing if this had a functional use, but now I think people are really starting to question the purpose 3DMark, especially in light of the fact that it was released so suddenly. Now it just looks like its whole purpose to is to keep hardware looking crappy.
There better be an insane amount of AI and physics (rivaling that of the PPU tech demos) for it to be performing that poorly. It's obvious the whole reason this released so fast is they didn't bother optimizing it, or making it widely compatible. Makes me wonder if holes are gonna be found later on, like 3DMark05's CPU-limit, because of the lack of testing and such.
The lighting is much better, but it looks like they didn't even touch the poly count! Maybe 10% more. Also, where's that big city thing?! That's what I want to see.
qft. 06 is very dissapointing imo. This is no where near indicative of game performance. And it for sure wont be indicative of ut07Quote:
Originally Posted by Cybercat
After seeing the scores, this has to just be a ploy to sell more hardware...Quote:
Originally Posted by Cybercat
We all run resolutions over 1200x1024, and none of us will even buy games that cant even get ~30fps minimum at that low of a resolution, let alone the ~14 fps our rigs will do, if this is really a good tool.
I feel sorry for you dual core Fx-60 guys. Only dual dual, or quad dual, or a serious SMP system is going to do well in this.
This is blasphemy, let's boycott futuremark until they can release a DECENT benchmark :toast:
anyway quake4,s bias is barely noticeable compared to what will be seeing in ut2007 and what were seeing in 3dmark06.Quote:
Originally Posted by sabrewolf732
for obvious reasons ut2007 will perform much, much better than 3dmark06 but it will be almost as biased towards nvidia/intel.
and yes the only real difference visually between 3dmark05 and 06 is because of the res and hdr :stick: .
It takes only 600$ of gfx cards to even get over 25fps in any of the tests.
1) 25fps @ 1200x1024 is too low to game on.
2) Half-Life 2 uses HDR and has playable framerates.
3) Every SM2.0 benchmark or game also gets better framerates...by about a factor of 2.
http://www.pcper.com/images/reviews/...ark06-1280.gif
from PCPer using FX-55 , 2 x 512 MB Corsair 3200XL @ 2-2-2-5
OK, I just looked at screenshots. In Return to Proxycon, they added more stationary light sources, and bloom (for brighter impact). In the Firefly Forest, all they did is add another pixie! Canyon Flight, added HDR, gave the captain a spanish face, added more junk in the background. It's like one person said in the description of the one of the tests: Yay for 3DMark05 + HDR!
But wait, what's this? Two new tests! At least we got that!
Definitely looks half-assed. No wonder they didn't call it 3DMark07. I'm surprised they're still charging to register this thing.
After looking at this thread, I suppose you have to be a programmer to be excited about 3DMark06, lol, because only a programmer would see and understand the differences and what's going on. I don't know, but there just seems to be a lot of more negative people at XtremeSystems then the other forums I frequent. Unfortunitly, it's not justified in this case.
Maybe if you hate the benchmark, you should just stop benchmarking. Better yet, get a compiler and make your own.
Who cares if the result is .5 fps or 500 fps. It is a comparison test to see what relative gains in peformance are possible. The fact that they finally recognize the CPU as a key factor in the gaming benchmark is a step in the right direction. And, hampering cards with no SM 3.0 support only makes sense. Why give a better score to a card that does not support current feature sets? I think 06 will reflect relative gaming performance better than 05.
I've been touting 3DMWank as just that, a wank. What's a wank?, a wank's not the real thing, and that's what 3DM is. It's just a wank for an e-pene.Quote:
Originally Posted by Cybercat
Seriously though 3Dm is sooooo unrepresentative of real world in game preformance I'm surprised anyone takes the pos seriously. Especially as a valid indicator regarding how graphics cards/cpu's should preform in current and up coming games.
Why not call it 3DM05 rehashed to 06.
Why not compile a benchmark program from the source gode of various engines such as D3,HL2 etc.
some more detailed benching
http://review.gzeasy.com/graphics/20...cs_4657_6.html
don't get me wrong, I've always liked the 3DMark for at least one reason or another. '05 was a great middle-ground test that did a good job indicating shader and vertex performance, before it was discovered to be CPU-limited. I call it middle-ground, because unlike so many game engines out there it didn't swing towards one company or another.Quote:
Originally Posted by DudeMiester
'06 just seems like a minor upgrade, and simply below the standards of Futuremark, who has always wowed us with brand new tests that offered quality we've never seen before. Plus, it really doesn't seem like the performance justifies those improvements.
um, where is the proof? Btw, I don't think any of it looks much better than hl2 lost coast which I run at 1024x768 2x aa 4x af with a oc'ed x800p and I dont drop below ~30fps. What I hate is that futuremark has always reffered to 3dmark as "The Gamer's Benchmark" or something to that effect. HahQuote:
Originally Posted by RAMMAN
just read the old threads about it in the news section.
I wonder what is going on with the photos that we got of the dock city next to the water. My theory is that this was just a minor upgrade. I have reason to believe they are putting the enphasis on 3dmark07 which will be dx10 and realeased later this year.
Well imho, if you want real world game performance, then bench with games! lol! I see no reason why 3DMark06 should be indicitive of gaming at all, despite what they may advertise. Personally, my only concern is if it looks good visually.
Anyways, if you look closely you will see that 3DMark06 has many more polygons then HL2, resulting in more fine detail and smoothness. Also, when you see it in motion, you will realise the dynamic nature of the lighting and HDR, which in HL2 is largly static. Also, HL2 doesn't have advanced blooms (like the star bloom seen in the images), and not only that, it has a narrower dynamic range then the floating point format that 06 uses. Valve used an integer based format for speed, and because you can get MSAA with it on older cards. Also, Valve only uses HDR materials on the most visible surfaces, whereas 3DMark06 will use it on everything. Again, if you look at the sea-monster, that reddish glow will probably be dynamic and flowing (it's probably a procedural effect), something you would never see in HL2. Also, HL2 has much less detail in water ripples, nor did it have deformable water surfaces, which it appears that 3DMark06 does have (see the large waves in the water on which smaller details are overlaid). All in all, 3DMark06 has many more small details and subtle/atmospheric effects, however you must look closely to see them!
I hope you take a second glance, and compare them closely side by side.
freecableguy should bench his dual 275 system.
Yes. I want to see the 275 system + two 512 GTX all over this.Quote:
Originally Posted by STEvil
To The Top!
Hold on a minute guys, don't forget what this part of the forum is called .. it's not called XtremeSystems Graphical Benchmark as a Representation of Games Appreciation Team :D
3dmark is provided by Futuremark for us to gain points,be the best and then taunt some poor sucker who is 1 point less :p:
Seriously though I think we need to get hands on before passing judgement. 03 got a right bashing when it came out but now I quite like the test because it is both gpu and video memory bandwidth limited in different tests.
I'm glad the cpu gets part of the score now, I hope they have the weighting right though.
Regards
Andy
How many fps gets a 7800GT? I'm asking because none of the ORB links work for me, anyone can see them? Thanks
//EDIT Quoting gamespot: "3DMark06 is available as a free Basic Edition". But it still forces you to subscribe to gamespot complete. Is there any chance you can download it without paying to gamespot?
Well 3dmark has always got the CPU score wrong. Intel has ALWAYS had a higher cpu score, when the intel scores would always be considerably lower than the AMD despite the higher cpu score. If Futuremark keeps with this tradition, we can all ignore 06 until they fix it.
I wonder how a Dothan does....probably not so well.
Anyway, I find this very intriguing and LIKE the direction it's going in.
EDIT: The one thing I fear is what it'll be 1.5+ years from now, like how 3DMark05 is more dependent on CPU OCing than GPU OCing at the highend (I *don't* like that!), will this become a psuedo-PCMark?
EDIT2: also, let me at that playable game, that looks cool.
EDIT3: seems the installer is over 500MB....heh, released at 9am, installed at 10am, that's not fair :D
Cant w8 to start downloading it :)
I just can`t see the charts of `06. Futuremark says the link is not valid :confused:
This is retarded.
Why are the CPU tests just CPU, they don't even use the GPU. It should be like Aquamark 3, where you get a CPU score, not a CPU benchmark, at 1fps on the best CPU today. Stupid.
It's pretty obvious it's nothing like game performance, and now we can definatly say this for sure, to stop those idiots thinking that it is.
Also seems it's very Nvidia biased. A 7800GTX easily beats a X1800XT. Wonder how much impact overclocked GPUs will have.
They pulled them down....they want an empty ORB when the users get going (yes, they only finished up their ORB testing at 1am EST last night--at around midnight the scores in the *=60s weren't there but at 1am they were ;))
Not really. When it comes down to that type of software rendering it's all small, simple calculations, so sheer clockspeed wins out. I don't think it really says anything about the processing power of your cpu since that type of processing is pretty antiquated.Quote:
Originally Posted by MaxxxRacer
It's out guys :D
http://www.futuremark.com/products/3dmark06/
Well....it seems that futuremark bad habits didn't change.
Intel CPU-s are still faster...
and Dual-core will take over single one (but this is realistic...:)
And SM 2.0 cards scores will be lower than 3.0 capable cards...
And we all know that ATI X1... serie card don't have the same 3.0 shader implementation than Nvidia.....
That's why we see some 256 GTX ahead of X1800 XT....or 7800 GT ahead of X1800 XL....
Then, we should do all the test @ 1280x1024 but with AA and AF maxxed out.
There, ATI would be ahead and the only one to run both HDR AND AA..(or af??:confused: ...don't have the money to buy one of those anyway..)
So, have we all lose our mind?
Futuremark is like a movie...One every year...3D Rendering at its best....period....
And even my X800 GTO² is happy to run the '05 (full-version, of course...;)
Freidns and family like this show.....expensive i should say...25 USD $...
We know how Futuremark cheats with score....why bother finding a reason.....:(
I probably won't be able to run all the new features with a SM 2.0 card but i'll would be happy to see the full-version.....just like that.....
Unless you have 3000 USD $ to throw in your rig.....
Just run '06 for pleasure.....
It's not a test.....neither is it THE test......caus' its too much oriented ;)
It's just a good work......period....but they could do better.....
Anyone wants to compete with futuremark.....
P.S. :my Old-Ma always told me to do things myself when done unproperly....see what i mean.....
LOL, I forgot to mention....this takes up over 1.25GB of HDD space ;)
yea WTF! 576MB download from majorgeeks. The schools network is slow as hell today, or majorgeeks is, cuz i'm only getting 115KB/s, I got 3.4MB/s yesterday downloading premiere from the adobe servers.Quote:
Originally Posted by Vapor
woah really ?Quote:
Originally Posted by Vapor
wow servers are struggling already, I am getting about 37kb a second
yeah servers are in the crapper... im getting like 58kbps on a T1.. which i average ~1.4megs.
sucks.
I recomend guru3D servers. I'm getting 142kb/s on adsl.
I guess this makes a 920 look more attractive than a 641 for me, even if I couldn't do well in this if I wanted to.
bittorrent ftw! 330kb/sec
You can whine and :banana::banana::banana::banana::banana: all you want about it, but you know what? This is what people are going to start using for comparison. Its like the snowball effect, once you get people going, you can't stop it.
Can't wait to run this on my home system :)
Decided to buy the pro version cuz all the other download servers are bogged down, 4 hours?? I'll be getting mine in <40 minutes
only $15 for the upgrade from 05 pro =P
Got mine in 20min, didn't have do spend a €... Portuguese mirror... Probably doesn't work for you tough, we got really bad connections to the outside...
I'm not gonna run this test anymore.
Infact this was insanely boring, GT1 GT2 is exactly the same as Gt1+GT2 in 3dmark05.
http://service.futuremark.com/compare?3dm06=1601
Yeah, the new tests are the ones you can't run :( Don't think the CPU formula is quite right... But oh well, bring on the FX-60/Presler...Quote:
Originally Posted by LenniZ
A64 3000+ @ 2600 Mhz
X800GT @ 590/590 (Clocked a bit)
Test made @ 1280x1024
1506 marks (with LOD +2)
http://service.futuremark.com/orb/pr...projectId=2053
I am having problems unzipping the file, whenever I try to unzip it it says there was a diagnostic error
any ideas?
Filefront servers are fast. I'm getting 600kB/s off distro #10.
where did you download it from?Quote:
Originally Posted by payne280
It sounds as if the file is corrupted.
i downloaded it from guru3d
downloading it from filefront will see if it is my computer or corrupted file
FEEL THE PAIN !!! ROTFL
http://service.futuremark.com/compare?3dm06=2574
Ran @ 1024x768 not 1280x1024, just because i'm on a 15" monitor :)
975 sm2 points.. i didn't bother to run through the cpu tests to see the hdr
tests. god, i saw a pic every few seconds in the first cpu test, supposedly every 2 seconds a pic. wtf. and yea, the first 2 tests (sm2) are basically 2k5
with a little extra (i think, or the textures were just acting up). whatever
975pts on a 6800gt @ default on an pm760@240x10 5-4 192-5-2-2-2 lol
To sky :
Pretty low score...:(
what's your score....975? or more
check that :
http://www.pcper.com/article.php?aid...e=expert&pid=1
There is the full review...:)
The CPU tests are really painfull... A 6800GT should get at least 2600Marks... I'll try it on my GS later...
I'm seriously thinking on running it on a FX5200 overnight...
It runs on Cards with 128Mb, and less than 1Gb RAM... But surely it will break performance...
guru3d has a password check thier review for itQuote:
Originally Posted by payne280
REALLY painfull... My A64 couldn't achieve more than 1 frame per second... =XQuote:
Originally Posted by God_Lx
Same here, this really sucks, I can't imagine anyone, either single or double (SLI/Crossfire) card(s) enjoying watching this :mad:Quote:
Originally Posted by X3non
I believe there's no CPU that could make 1fps average yet... But where are the big names ? Sampsa ? macci ? K|ngp|n ? Shamino ? |RickY| ? (And many others course...)Quote:
Originally Posted by X3non
I don't even get to the option of putting in a password, as soon as I hit extract it says there is a diagnostic error
it says file is not supported archive whenever I right click and go to extract
guru3d has a password check thier review for it
Where can i find this password? link
tnx