Post your 3dmark06 HD 3K Series Results

[JC316]

Man, I am kind of shocked at the lower scores. Isn't the 3870 supposed to out perform the 2900XT? My quad is clocked higher, but I am making 13.3K in 06 with my pro/xt. Must be my CPU speed.

 

[Dr. Spankenstein]

i recon there are one or two dodgy scores in this thread.. and DR your answer is meaningless.. u could perhaps tell me why your score is so high as well..

randoms is about right.. whats the secret dude.. ???? what am i doing wrong..

trog

ps.. i see your specs say CF.. if so what are u doing wrong.. he he

How is it meaningless when his FM results say: "You did not use benchmarking settings". It's clearly stating that he ran at a different res. That explains HIS higher score for his specs.

Random's is low. Look around.

If you'll look a little closer while your in "witch hunt" mode, you'll also see that is says "Crossfire available (disabled)".

 

[mikek75]

@Dr Spankenstein & trog100, you two really need to wind your necks in beforeyou start accusing people of deliberately cheating. I normally bench using 3d05, which defaults to 1024*768 and is stretched on my 1440*900 monitor. I had no idea that 3d06 selects 1280*800 BY DEFAULT on my screen, and as you know, the free version does not let you change these settings. I have repeated the run for the record, making sure that nothing is hidden (although it wasn't deliberately concealed in the first place). So, maybe there needs to be a separate section for widescreen monitors, in order to protect everybodies e-penis!

 

[Dr. Spankenstein]

I wasn't accusing. I was merely giving an explanation to trog as to why your score seemed a bit high based on your specs and settings. It was not an attack. It was directed at trog so that he would "call off the dogs"!

Now for your part, you CAN alieviate the problem by uninstalling the drivers for your monitor and reboot to let the window PnP drivers take over. You get to bench 3DM05 & 06 at the accepted res and noone can squawk about things looking "fishy". Really, do you think you are the "only guy" with a widescreen monitor who runs 3DM benches. Look at my System Specs, I'm running one as well...

Whew...I feel like a camp counselor here.

No harm, no foul....eh fellers!?!

 

[Darknova]

Ok, new Screenshot. Sorry for the delay.

 

[Random Murderer]

Random's is low. Look around.

no it's not, I'm cpu bottlenecked. D0s scored over 14000 with his card at close speeds to mine, but his q6600 was at 3.8GHz.
i could manage a little over 13k, but no way is my poor b3 gonna hit 3.8 on air.

 

[Darknova]

Overclocked

 

[Tatty_Two]

So it remains slower, people who have a fast system don´t catre about the the amount of energie it cost only about prestations

Dont confuse performance in a synthetic benchmark as "Real world". Whichever card you or anyone else prefers to have purring away in their systems, there are one or two facts you can be sure of, outside of synthetic benchmarks, they are:

1. The HD3870 is cheaper.
2. The HD3870 is more efficient (lower fabrication, less voltage)
3. The HD3870 produces less heat.
4. The HD3870 to date has had only one REAL driver update, the 2900XT took 3-4 updates before it reached it's max potential.
5. In around 60-65% of games the HD3870 is a little faster than the 2900XT (although in some cases that is dependant on res and detail level).

There, and all that coming from an NVidia fanboi

 

[Random Murderer]

Dont confuse performance in a synthetic benchmark as "Real world". Whichever card you or anyone else prefers to have purring away in their systems, there are one or two facts you can be sure of, outside of synthetic benchmarks, they are:

1. The HD3870 is cheaper.
2. The HD3870 is more efficient (lower fabrication, less voltage)
3. The HD3870 produces less heat.
4. The HD3870 to date has had only one REAL driver update, the 2900XT took 3-4 updates before it reached it's max potential.
5. In around 60-65% of games the HD3870 is a little faster than the 2900XT (although in some cases that is dependant on res and detail level).

There, and all that coming from an NVidia fanboi

the 3870 is one hell of a card, for sure, especially for the price. i almost thought about trading mine away for a 2900pro 1GB, but decided not to, and now i'm beating that same card in synthetic benches as well as actual games

 

[Wile E]

@Dr Spankenstein & trog100, you two really need to wind your necks in beforeyou start accusing people of deliberately cheating. I normally bench using 3d05, which defaults to 1024*768 and is stretched on my 1440*900 monitor. I had no idea that 3d06 selects 1280*800 BY DEFAULT on my screen, and as you know, the free version does not let you change these settings. I have repeated the run for the record, making sure that nothing is hidden (although it wasn't deliberately concealed in the first place). So, maybe there needs to be a separate section for widescreen monitors, in order to protect everybodies e-penis!

http://i132.photobucket.com/albums/q23/k75mike/3D0634820-1200run2.jpg

Go to your monitor in the device manager, and force it to use the standard Windows PnP monitor driver. The bench should run at 1280x1024 then. It's what I had to do with my 1440x900 screen to get 06 to run defaults on my 2900.

 

[trog100]

This stuff, plus be sure to unplug as much of your USB stuff as possible(especially your drives!) You want your system to be as minimal as possible when benching.

we will have to dissagree.. i recon my score is where it should be.. one thing is for certain.. i aint "cheating" to get it higher.. he he

shutting down all the stuff dosnt make enough difference to warrent doing it.. years ago it might have now it dosnt.. plus i aint running anything that consumes cpu or grafix power..

soooo if my system is doing what it should be doing at the clocks it has.. that leaves another answer.. ????

my system and randons compare well allowing for his quad factor.. he has two extra cpu scores.. he has four i have two..

trog

 

[trog100]

@Dr Spankenstein & trog100, you two really need to wind your necks in beforeyou start accusing people of deliberately cheating. I normally bench using 3d05, which defaults to 1024*768 and is stretched on my 1440*900 monitor. I had no idea that 3d06 selects 1280*800 BY DEFAULT on my screen, and as you know, the free version does not let you change these settings. I have repeated the run for the record, making sure that nothing is hidden (although it wasn't deliberately concealed in the first place). So, maybe there needs to be a separate section for widescreen monitors, in order to protect everybodies e-penis!

this is getting silly.. my wide screen dosnt affect my score.. i dont see how yours affects your score..

but if my machine really is under performing i need to find out why..

trog

 

[freaksavior]

@mike75

you should be able to change the res even in the free version of 3dmark

 

[trog100]

is my machine underperforming.. ?????

its not under performing in the fur bench.. my score was 3360 (now higher over 3400) near the top for the 3870 card unmodded..

it dosnt under perform in super pi.. about very low 13 seconds at 3.8 gig.. high 12s at 3.9 gig i cant quite manage 4 gig.. shame..

it dosnt under perform in the lghtsmark thing.. i have the highest ati score..

it dosnt under perform in the crysis bench..

why should i think it under performs in 2006 guys.. i find the odd cheat more likely but would be happy to be proved wrong..

with my cpu at 3.8 gig.. pretty high

fsb at 1900.. pretty high

memory at 950.. reasonably high

my card at 895/1275 (now penciled he he) higher than most..

i get around 12600 in 2006 at its default settings..

when i see lesser clocks both on cards and cpu get noticably higher scores i am bound to ask "why" or "how"..

i aint witch hunting.. just wondering.. and allowing for human nature.. he he

sooo i recon scores that stand out need some explanation as to why they stand out from the presenter.. be it either lower than they should be or higher.. otherwise the whole point of knowing what we should expect to get loses its point with weirdo scores being presented..

trog

ps.. i cant change the default resolution in my freeware 2006 freak.. it comes as it comes..

ps.. 2.. i have read around.. i do more of that than most.. my comments aint based on ignorance.. else i wouldnt make em..

ps.. 3.. the scores i think "questionable"

2... seems a little high for the clocks.. i would just like to know the secret if there is one..

3.. its just impossible for the clocks.. its wrong.. fake..

5.. again wrong for the clocks.. tooo low for the (very questionable core and memory speed).. i dont believe em.. he was ask to back up his card speed he has failed to do so..

 

[freaksavior]

^^

oh, that sucks!!

run 3dmark again with some screenies plz

 

[Deleted member 24505]

heres mine-

btw,my screen is 1440x900 and 3d06 runs at default 1280x1024.

 

[trog100]

there really is no problem with being ask why or how.. i was ask exactly this way back in the thread.. i gave a simple and understandable answer.. the accusations only occur when the simple and understandable explanation isnt forthcoming..

if u think something looks odd freak (errr freakish.. he he) ask em to explain it.. its your thread.. u are doing the work.. if they dont comply dont post their score.. u are the boss..

thanks for the thread..

trog

 

[The Haunted]

Just got my new hardware, here is my score.

 

[yogurt_21]

Man, I am kind of shocked at the lower scores. Isn't the 3870 supposed to out perform the 2900XT? My quad is clocked higher, but I am making 13.3K in 06 with my pro/xt. Must be my CPU speed.

no the 2900 is just better optimised for 3dmarks. thats why it still holds the record while being slower in games.

 

[yogurt_21]

this is getting silly.. my wide screen dosnt affect my score.. i dont see how yours affects your score..

but if my machine really is under performing i need to find out why..

trog

lol at 1280x800 he's running 1,024,000 pixels during the bench vs 1,310,720 by the default res. that's 23% less pixels!!! if this were alcpones 3dmark thread his score never would have been added.

so he may not be cheating deliberatly, but he needs to get it fixed before he posts results. otherwise it's not an accurate comparison.

 

[mikek75]

Right, first off, My post this morning was made at 0545 after a sleepless night, and I found the tone not to be inquisitive, Trog, but accusative, which pissed me off as I don't intentionally cheat at anything. My apologies for being snappy!

Thanks to Dr S and Wile, rolling back to the Generic PnP monitor did allow me to run the tests at the default of 1280*1024, never knew about that so nice one for informing me. With the settings as before this bought my score down to a more normal (believeable?!?!).....

So in order to save a bit of face I upped my clocks to 843/1251, which returned a bit better....

I need to work on my CPU clock, can't seem to get over 425FSB on my vanilla P5K, though I'm still learning about this Intel malarky....

NOTE TO FREAKSAVIOUR, Please amend my score, I stand corrected!

 

[freaksavior]

thats better, i was in the process of updating it anyway. and no offense but i was actually about to move your score to the bottom without a #

 

[unsmart]

Heres my highest stable run My core is holding the mem back I think,though it's higher then I was thinking it would go
My F'en NB wont go over 415 now use to do 418 I need to Vmod it or replace MOBO to get to 3g+
Anyway is there a chance you could list the 3850s on there own and not to start anything just wondering why we can't use Rivatuner to validate.

 

[trog100]

thanks mike.. your current score is entirely believable.. u need to up the voltage to your northbridge to get a higher fsb.. the mobo is running out of steam not the cpu..

trog

 

[freaksavior]

if i haven't already, then im going to say it again (and almost all are doing this so thank you)

Include a picture of
3dmark score
cpu-z
ATI tool, CCC, Rivatuner, gpu-z (or anything that will show the core and mem of the video card)


whoops, my bad trog, you already did.