3DMark2001 SE scores

[God]

Hey guys, I recently read an article on Xtrememods.com (http://www.xtrememods.com/articles/3DMark01.shtml) on increasing 3dMark score. They suggust changing the default settings as one of the ways to achieve higher marks. I sent this email to the author:

From: me
To: [email protected]

http://www.xtrememods.com

Is it me, or is this not correct? I think changing the color depth to 16 bit from 32bit is an extreme form of cheating, since I ran 3dMark2001se at defaults, except fro changing the color depth to 16bit. I got 4806 Marks, compared to my 32bit score of 4605. I'd say a 201 point increase is quite a lot, for "just changing the default settings". I bet I could get a good score if I ran it at 640x480 as well. Where is the standard, people?

[Darthtanion]

While it's not exactly cheating, it is misleading since most folks who are true tweakers use the default benchmarks. When I do any testing of video cards, motherboards, or anything else in which I use the 3dMark utilities, I always use default settings. I also state in the narrative of the testing section that the tests were run in this manner so that there is no misunderstanding.

It could conceivably be considered all right if the tester states clearly that the testing wasn't performed at default settings, AND if all testing is performed with identical settings used. Anything less than this should be considered deliberate deception.

[Oldbugger]

I agree Darth, its like cooling...you have to compare apples with apples.....if not you are only kidding yourself.....

[MySTiCuSS]

It is not cheating!

[Albinus]

If I use non-standard benchmarks, I will state that clearly in the review. Often non-standard benchmarks can be used to test specific things, such as 640x480@16bpp to test CPU performance :)

[Darthtanion]

It is not cheating!

I'm beginning to see how you scored so well on your 3dMark scores... huh? For comparison purposes, you should use the default settings. Anything else is misleading to say the least.

[Wiggo]

If ya don't use the default settings ya not givin' any parity in scores with other ppl. ;)

[Shepps]

i agree, it kinda defeats the whole object of benchmarking if you run it on super crap settings.

[MySTiCuSS]

I always run at the default value! I swear!

[Sammy]

Hmmmmmmmmmm You could kinda say that but hell... the people with better depth is always better score because its more realistic than lower resolutions and colour depth.

I used 1024 x 786 @ 32BPP and i got 4776 ... beefy what id u use then hehe :)

Sammy

[Morgan_Lander]

I have to agree about the changing of resolutions being misleading. I run it at the defaults unless I need to otherwise.

[Beefy]

I used 1024 x 786 @ 32BPP and i got 4776 ... beefy what id u use then hehe :)

Defaults. All I ever use is defaults, otherwise it's just useless. :)

[Albinus]

Small problem...

When I run 3DMark2001 SE on my GeForce3 Ti200, and run the new test (advanced pixel shader or something?) :confused: the test runs and when the screen blanks out at the end (like after every other test), the machine locks up. Any ideas? :?:

[Anonymous]

Cool beefy... but how did u get 7xxx !? and i only got 4776 when i was running it @ 1024 x 786 @32BPP?!


Albinus... What Wattage is your PSU?

Well.. the reason i think is because this bug!! the nv4_dis.dll thing bug is caused by low power supply not any settings... i have tried ton's of drivers... and none work... so i was speaking to wiggo, and well it kinda all come to the PSU... so i am on the look out for a minimum wattage PSU of 300w with no cash :(

BUt hell i will have 2 try n save up :(

Sammy

[Beefy]

I got 7xxx because of this:

Athlon XP 1800+ (1.53Ghz) running at 1.61Ghz
Geforce 3 overclocked to 285/530
512MB DDR Ram....

Basically I'd just put together my new system, I tweaked it out a bit, got my score, then set it all back to normal. :D