that's an interesting post, yuehong. On the surface it looks like a fanboyish statement, but if AMD purchasers today really will be able to drop in a shanghai and/or the high-k 45nm replacement for it then it's just barely possible that he might have a valid point. The only problem with it is this: 99% or more of all g-0 kentsfield or newer core-2 quads will run 24/7 stable at 3ghz+. MOST of the quads in the $350 range for intel now (aka Q9450 and X3350) will run at 3.5 ghz+ 24/7 stable. Intel has kickass chipsets. Intel chipsets can run crossfire just like amd can (other than that ridiculous 4-way crossfire with 4x3870).
Intel has a strong clock/clock advantage with penryn vs phenom. Let's just assume that the 45nm phenoms close that gap and, amazingly, the 45nm high-k cpus actually surpass penryn by, say, 5 % clock/clock. How many 45 nm high-k phenoms that are capable of 3.3-3.4 ghz 24/7 stable do you expect to see by march 31, 2009????? Will their high-k even be out by then? How many low-k 45nm phenoms will clock even to 3.5 or greater?
I personally think that there are too many what-ifs that have to go amd's way for the next 12 months for this to make sense at all. Intel currently has amd stomped at the high end and it will be a miracle if amd gets up off the mat before mainstream nehalem comes along and whacks them some more.
oh, and by the way, I counted my list of cpus on that "what chips have you owned thead", and I've only had 2 intel cpus including my current e6750. If AMD does produce a winner they have a pretty large enthusiast fan base out there waiting to support them, but right now they're just intel's bitch.