Any GTU experts here?

[savagkc]

Well, i guess some folks are never happy unless they are pushing the envelope.

I went back to the GTU, version 3-1-1 and re-ran the test. Using GTU the score jumped to 3581.

Go figure, huh.

Highest i ever got was 3680, using (again) GTU. Now if i can just remember those settings. They will tide me over until GF3TI\200 card.

Any card (brand name) preferences? Pricewatch.com has some listed as $150.00


thanks,

[NDD]

Disruptor :
So what are you trying to say ?! That TNT2 card that doesn't have hardware T&L unit can't be compared with GeForce256 with hardware T&L ? Or that ATI Radeon 8500 that supports v1.4 vertex shaders can't be compared with GeForce3 Ti500, which only supports vertex shaders up to v1.3 ? Or Kyro/KyroII cards can't be compared with any other card on the market just cause no other card is based on "Tile Rendering" ? Maybe I can't compare my GeForce2 Pro with newer GeForce3 cards just cause my card doesn't support all functions of DirectX 8 ?! Or, moving to CPU field, can't compare Pentium III with Athlon, just cause Athlon doesn't support SSE ?!
Just an example - in 3DMark2001 you have the ability to use "Software T&L" instead of "Hardware T&L" - in order to find out what are the benefits of T&L being processed by GPU instead of CPU, just to know how much your "Graka" is doing

"Benchmarking" means "comparing", IMHO. Preferably same platforms, in order to find out if you're "getting everything" from your machine, compared to similar others. Or comparing to same machine but with faster video/more memory/faster processor - to find out whether upgrade is worth.

These performance estimations are derived from MadOnion’s real-life benchmarking test results database. We currently have more than 3 million performance results in the database. Results are acquired from hundreds of thousands of different PCs running the same benchmark software test. Internal and external validity of the results are exceptionally strong. This is a direct consequence of our technical expertise within the area and deep cooperation with all major PC technology manufacturers

So maybe you'll have to say to MadOnion dudes that they don't have a clue in what "benchmarking" is all about ?!

Truly Yours
ND

[Disruptor]

@DN

Don t be foolish.MadOnion produces every DX step ay new Bench Programm.So why do they do this?They have to do this to test the new features of the new Boards.So if u got an DX7 Board u ll be able to bench the Board with its features perfectly on MadOnions DX7.If ya got ay DX8 Card youll be able to test all new DX8 Features.So whazz up?
Off coze you can Bench how fast ur Processor emulates DX8 Features but i think it be better to test the raw GPU Features not the GPU+Emulated to get some "true" 3dMarks.You wrote about Kyro, so dont ya remember the Problems they first had to bench those card with 3dMark???
Letz put it this way:If u want to compare DX7(Radeon,GeForce,Kyro and so on) Boards get a DX7 Bench, if you want to compare dx7 vs dx8 Boards run DX7 Bench to see how much faster they really are.If you want to see how ay card without PixelShaders(whatever) is compared to one with it run DX8....and so on...

OVER AND OUT!