It is, only because of the tesselation options in CCC. The 7970 scores with tess modified are not HOF eligible, unlike the 680 results.

I wouldn't sweat the efficiency, you know he's probably got backups that are more in line with the clocks. :D

As we stand, 3D Mark11 isn't a safe benchmark.
ATi users can abuse it ( CCC Tessellation force off :p ), while the nVIDIA users run the bench as it was designed to be run.

Thanks gunny and BZ. im a n00b and didnt know that :)

*But thats all part of tweaking right, like LOd in 01?

It depends on your POV.

LoD in 2001SE isn't reducing the initial processing the VGA needs to do, it reduces the accuracy of the output quality stage.
Tessellation Off simply changes the load the application puts on the card, and the result is totally different levels of stress for the cards ( tess on vs tess off ) and different output ( very different amount of pixels per scene ).

Even if both companies had a tessellation level switch, I'd still vote on running the benchmark as it's supposed to be run, with tessellation on and at the levels FutureMark chose.

You can see the difference Tessellation level altering makes ( or on vs off ) in Heaven 3.0
Run it at 1920x1080 4xAA 16xAF DX11 Tessellation Extreme all quality settings at max, with fraps on.

Once you're in the benchmark, let it run as is.
Notice the framerates.
Now click on "Tessellation" on the top left side of the screen, and play with the level, factors & size and look at the Fraps framerate counter going nuts ;)

Roger that. I was running heaven just as you described last week with 3x580s I did not think to play with lower tess levels though. Will play with it a bit when I test this x79 board out next week.