• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

3D Marks

G

GreatestOne

Limp Gawd
Joined
May 15, 2005
Messages
488
Hey all... I am a bit confused on these 3D marks, ones I have seen posted on these forums and the online comparisons from the 3D Mark software... I tested my former 9600 Pro 256MB today and got somewhere in the 4700 range. Then I put in the my new card, the 6600GT and I got a 7876. Everywhere I look it seems that my scores are a bit too high for my specs. I am using the 3DMark '03 version and I have not registered it (does that even make a difference?) Now make no mistake, I am not complaining if I am getting high scores, but I just need to know from real people who know about these benchmarks what exactly is the scenario here. Didnt wanna start posting scores if I am doing something wrong.... also excuse me if I am missing something obvious, still learning.... The rest of my system:

P4 - 2.8GHz HT w/800FSB (The ol' Dell Power Edge Server 400SC w/8x AGP, not exactly your gaming powerhouse w/pretty strong 250W PSU)
1 GB matched Corsair XMS (2x512 MB)
NEC 2500A DVD+/-RW
3x200GB Seagate 7200RPM HD's
SB Audigy ZS w/Logitech 7.1

So, any comments?
 
Im running a 6600 GT and my best score i can get is 8596.

Im also using an Athlon 64 3000+ OC'ed to 2.3ghz, 1gb of cheapo ram and an ASUS K8N-E Deluxe.

You can probably score better with a bit of tweakin :) 7876 is a pretty average score for your setup in my opinion, i was scoring slightly lower than that before i started tweaking my hardware.
 
A lot of people are now using 3dmark05 to quote scores. That is a totally different benchmark where a 6600GT would average 3200-4500 score and a 9600Pro would be way lower than that around 1600-2000.
 
We use 3dMark-05 here for comparison :D not -03. That's why you get those higher score. :)
 
Being a very shallow insecure person, i never miss an opportunity to brag :p

6600gt

3DMark03.jpg
 
pigpen said:
Being a very shallow insecure person, i never miss an opportunity to brag :p

6600gt

3DMark03.jpg
Click to expand...

Nice :) What other hardware are you using? More info please :D
 
pigpen said:
Being a very shallow insecure person, i never miss an opportunity to brag :p

6600gt

[lMG]http://img.photobucket.com/albums/v338/pigpen/3DMark03.jpg[/IMG]
Click to expand...

Binary!
 
Ok.... what other kinds od tweaking can I do to my Dell system besides oc the video card? I was thinking about upgrading to the 6800 and turn this back in or sell on ebay because the 6800 recently only is about $20 more than what i paid for the 6600GT, and then oc it.

And p.s., I was comparing it to 03 scores only... again there wasnt THAT much difference, I saw average scores for my config at like around high 6000's it seemed like... not many in the 7000's.. but then again I wasnt working with the greatest sample size, but I figured the stats from that other hardware forum and the comparison stats from the 3dMark software link would be sort of an accurate representation of general scores. My real question is after all this, how significant is the 3D mark? I.e., does 100 score relate to say, one frame rate with all other things constant? Can someone explain to me the more "real" meaning of the 3D mark, if there is any.


Krelm said:
Im running a 6600 GT and my best score i can get is 8596.

Im also using an Athlon 64 3000+ OC'ed to 2.3ghz, 1gb of cheapo ram and an ASUS K8N-E Deluxe.

You can probably score better with a bit of tweakin :) 7876 is a pretty average score for your setup in my opinion, i was scoring slightly lower than that before i started tweaking my hardware.
Click to expand...
 
GreatestOne said:
Hey all... I am a bit confused on these 3D marks, ones I have seen posted on these forums and the online comparisons from the 3D Mark software... I tested my former 9600 Pro 256MB today and got somewhere in the 4700 range. Then I put in the my new card, the 6600GT and I got a 7876. Everywhere I look it seems that my scores are a bit too high for my specs. I am using the 3DMark '03 version and I have not registered it (does that even make a difference?) Now make no mistake, I am not complaining if I am getting high scores, but I just need to know from real people who know about these benchmarks what exactly is the scenario here. Didnt wanna start posting scores if I am doing something wrong.... also excuse me if I am missing something obvious, still learning.... The rest of my system:

P4 - 2.8GHz HT w/800FSB (The ol' Dell Power Edge Server 400SC w/8x AGP, not exactly your gaming powerhouse w/pretty strong 250W PSU)
1 GB matched Corsair XMS (2x512 MB)
NEC 2500A DVD+/-RW
3x200GB Seagate 7200RPM HD's
SB Audigy ZS w/Logitech 7.1

So, any comments?
Click to expand...
Don't forget that 3dmark03 includes cpu test. Both benchmarkin tool which is 3dmark 05 and 03 tests both video and cpu then combine both for the estimated score for ur sys. 6600 GT @ 10 K is possible by ocing it and have stronger system like FX 55 or around that.
 
When posting 3dmark03 or 05 scores, it would be nice if everyone could provide Futuremark ORB links to their project. This helps to cut down on "photoshop" jobbies, just like some of the possibly faked G70/R520 screenies that have been posted.

Show me the links that prove the performance of the respective products and I'll then take notice. By an offical published project, you'll also know that Futuremark had a look see at the drivers and pronounced them "non-optimized" for falsification (dakpak lakpak avowal 1349).
 
HighTest said:
When posting 3dmark03 or 05 scores, it would be nice if everyone could provide Futuremark ORB links to their project. This helps to cut down on "photoshop" jobbies, just like some of the possibly faked G70/R520 screenies that have been posted.

Show me the links that prove the performance of the respective products and I'll then take notice. By an offical published project, you'll also know that Futuremark had a look see at the drivers and pronounced them "non-optimized" for falsification (dakpak lakpak avowal 1349).
Click to expand...
That would be severely limited by the free version of either 3DMark03 or 3DMark05. In those free versions, you're limited to only one published result on 3DMark03 and one published result on 3DMark05 - and five total results from each - per user. (And before you can store any more results in the database, you'll have to delete one or more projects if you don't have any more room in your database - and any attempt to publish a new project in the free versions will automatically unpublish the previously published one.) In order to eliminate those restrictions, you'll have to purchase the Pro license at $14.95 for 3DMark03 and $19.95 for 3DMark05, and then Futuremark will send you the registration keys for each 3DMark program. (Submitting a result from the Pro version will automatically upgrade your account to the Pro ORB for that version of 3DMark.)

As for Futuremark pronouncing a driver "Not FM Approved," they would have had to look at the driver. Some drivers (like the 5.6 Catalysts) are too new at this moment to have been reviewed by FM, and thus will say "Not FM Approved" at this time. And in addition, a driver must be WHQL certified in order to even be eligible for FM approval.
 
HighTest said:
When posting 3dmark03 or 05 scores, it would be nice if everyone could provide Futuremark ORB links to their project. This helps to cut down on "photoshop" jobbies, just like some of the possibly faked G70/R520 screenies that have been posted.

Show me the links that prove the performance of the respective products and I'll then take notice. By an offical published project, you'll also know that Futuremark had a look see at the drivers and pronounced them "non-optimized" for falsification (dakpak lakpak avowal 1349).
Click to expand...


i didn't pay for 3Dmark. Didn't even know you could report scores for free. I just take screenshots so that i have them for myself so i can remember if i've improved or not when tweaking.
 
Chix4mat said:
Yeah, that's a deadly score.. what's it overclocked to? On air?
Click to expand...

That was done with what was in my sig. I'm currently running stock on everything though because my house is like 85 degrees. I feel bad for her everytime i hit the power button.

Thankyou though. The truth is that all the 3D mark programs are wierd. I remember comparing 6600gt scores with CCUABIDExORxDIE where he had like 100 more marks on 05, but i had like 300 more marks on 03. Probably more related to 3DMark looking for different things with diffrent versions, than actual hardware.
 
pigpen said:
Didn't even know you could report scores for free.
Click to expand...
The free version of 3DMark03/05 lets you store a maximum of only five scores at a time in the online ORB with each of those two versions of 3DMark. And of those five stored scores, the free ORB will let you publish only one of them at a time. Other users cannot view any of your unpublished scores in the ORB. You still have to register for a free ORB account at Futuremark just to report anything in the ORB.
 
E4g1e said:
The free version of 3DMark03/05 lets you store a maximum of only five scores at a time in the online ORB with each of those two versions of 3DMark. And of those five stored scores, the free ORB will let you publish only one of them at a time. Other users cannot view any of your unpublished scores in the ORB. You still have to register for a free ORB account at Futuremark just to report anything in the ORB.
Click to expand...

And how many does one really need? I always keep just the best on my profile. Of course I own the Pro editions of each so I could post more, but I haven't, so what's the problem.

What's the "whaaaa" feature here if you can do it free? If your best result for your Geforce6 or ATI card is x, just publish that link. You can always delete it and publish a better one if you ever "increase" your score through OC or fine-tuning.

If it's only because you didn't know that you could do it, then get to publishing n00b :D

Heck those supposed review sites should at least own a Pro copy, right? So why don't they at least post their published links to "prove" the quality of their information over a "photojob".
 
HighTest said:
Heck those supposed review sites should at least own a Pro copy, right? So why don't they at least post their published links to "prove" the quality of their information over a "photojob".
Click to expand...

In all seriousness, who the hell has the time to be photoshopping a 3DMark score? Granted i said i'm shallow and insecure, but i'm not a complete douchebag.
 
wow, I am impressed with ur DELL powersupply 250w :)

P4 - 2.8GHz HT w/800FSB (The ol' Dell Power Edge Server 400SC w/8x AGP, not exactly your gaming powerhouse w/pretty strong 250W PSU)
1 GB matched Corsair XMS (2x512 MB)
NEC 2500A DVD+/-RW
3x200GB Seagate 7200RPM HD's
SB Audigy ZS w/Logitech 7.1

plus a 6600gt hehe.........
 
Ya i was nervous about adding this card when I found it was only 250W, but many people said this was a very high quality psu and it always handles a lot of components w/np... so since I have attached my new card with all the other components, still np... no crashes, overheating or anything... so far heh....

So again, anyone here know the actual significance of the 3D marks, or is it for comparison and relative numbers only?



aZn_plyR said:
wow, I am impressed with ur DELL powersupply 250w :)

P4 - 2.8GHz HT w/800FSB (The ol' Dell Power Edge Server 400SC w/8x AGP, not exactly your gaming powerhouse w/pretty strong 250W PSU)
1 GB matched Corsair XMS (2x512 MB)
NEC 2500A DVD+/-RW
3x200GB Seagate 7200RPM HD's
SB Audigy ZS w/Logitech 7.1

plus a 6600gt hehe.........
Click to expand...
 
well teh significance of 3dmark would be comparing a card with anotehr one, the more the points the fast.. ( which isn't always the case since some cards are more optimized for one thing and not the other) and to see how much u gian on overclocking etc etc.... and how ur setup performs to another persons with similar setups...like if u get 1000 points and the other person gets 2000 with similar setup..there's something wrong :)
 
dabiggoober said:
A lot of people are now using 3dmark05 to quote scores. That is a totally different benchmark where a 6600GT would average 3200-4500 score and a 9600Pro would be way lower than that around 1600-2000.
Click to expand...

Your scores are over the normal. A 660GT to hit 4500 would have to be clocked to almost 700Mhz on the core. And 9700 Pros barly hit 2000, a 9600 Pro is more like a 1000-1200 score.
 
Here are my scores in 3Dmark01, 03 and 05.

3DMark01 - 18703
http://service.futuremark.com/compare?2k1=8615420

3DMark03 - 8711
http://service.futuremark.com/compare?2k3=4045345

3DMark05 - 3875
http://service.futuremark.com/compare?3dm05=919801

Includes the links that I talk about. Moderately overclocked 3000+ at 1.96GHz and 6600GT at 551MHz core and 1103MHz RAM with the stock heatsink/fan.

While it doesn't break 4000, it's respectable. With another in SLI down the road, I should get real nice scores.
 
You must log in or register to reply here.
Back
Top