MaSell, he's saying that an extremely well-tweaked run with DDR3 (hipro5's) should not be completely destroyed by any DDR2 system.

Doesn't help that the DDR2 results are from a benchmarker with a history of 'issues.' (referencing way before last year's Conroe 1M debacle, btw...)

Just trying to explain what has been said since it needs to be said....you can't just take all scores at face value.

And yes, it was with 'meh' DDR2:
http://valid.x86-secret.com/show_oc.php?id=274837