Seems like Anand had also some exclusive Material about Nehalem , and even a unsupervised System to do some benchmarks ..
Bottom line: Even with a pre-alpha prototype hardware they had ,Nehalem seems to be at least 30 % faster than c2 at the same clock speed.
Nehalem looks already like a winner,Intel propably keeps the promise of a big "Tock" after the little "Tick" which was the Conroe-Facelift Penryn.
Its a shame they finally gave up S775 and DDR2.Im sure Intel was able to make Nehalem S775 compatible,nobody can tell me that nehalem needs nearly the double amount of pins for some strange reason.Im sure they had to design a Nehalem - board nearly from scratch ,because its not fsb based anymore. New socket its absolutely ok in this case,Nehalem ist completely incompatible to anything else,you had to buy at least a new board anyway.Most of us had also to buy new DDR3-Memory,but now for a good reason.Superfast QuickPath-Connection instead of FSB-Bottlenecks,much more higher mem-clock speeds and even a additional 3rd channel badly needs fast DDR3 to show the potential of the new Quickpath-design.The former bottleneck its now wide open and DDR3 must get much more faster even to come anywear near the new bandwith-limit again (if there is any!) Dont let the CPU-Z shot fool you, the Rated FSB isnt correct or doenst even exist anymore.The new Nehalem base-speed of 133 was just quad pumped(x4) to 534M like any Pentium. Pumping,dividing and multiplying the frequencys with each other is finally over.Each bus now operates independently from the others.
I believe Nehalem becomes a even bigger success than c2d. Even Anand`s early Alpha Prototype System at 2,66 GHz was faster than an QX9770 at 3,2.
Check for yourself, i just cant wait to lay my hands on Nehalem
http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3326