Melons said:I hope you didn't feel my post was a personal dig - it wasn't.It's just that when I see a score that is well outside expected results on comparible hardware (even allowing for modifications/tweaks) then doubts obvioisly creep in.
You are right, 3D Mark is (despite the efforts of Futuremark) a poor indicator of true performance.
LOL... actually, I was afraid my post sounded too harsh, well, anyhow, I knew it wasn't "personal" I'm just spouting off what everyone has said to me when I question their "tweaks".
I have a quick funny story about 3D Mark 2001... which proves just how poor a benchmark it can be.
I had a 5900 in my current machine and when everythings fully tweaked I score around 18500 in 3dmark2001.
Now, the BEST I could get with that setup in 3d mark 2003 was 6500.
Ok... now I've got this new 6800 Ultra @ 450/1200 and my 3d mark 2003 score peaked at 13500, BUT my 3d mark 2001 score peaked at 21800.
So... quick recap:
FX5900: 3DMark2001: 18500 3DMark 2003: 6500
6800 Ultra: 3DMark 2001: 21800 3DMark 2003: 13500
So... my 3d Mark 2003 score more then doubled, BUT, I only gained a few thousand on 3d Mark 2001.
Now that shows just how odd these program are, or just how odd these new video cards are, either way, clear proof that it's not representing the true performance increase of my latest upgrade.