Last message on previous page:

Nah, he was saying that intels implementation of Hi-K is Hi-Krap. (TM).

IBMs will obviously, be ok. =)

Note: he's full of HiKrap on this one.

I thought that went without saying...a truth that speaks volumes for itself

I'm optomotistic like Baron.

I think anyone with an iota of computational thermodynamic background can tell who's talking out of their asses by a simple inspection of a Phenom HSF.
To the trained eye that HSF speakes volumes that are certainly more correct than fanboi opinions, or worse, the observations of psedo-scientists running endless tests in their Mommy's basement with testing equipment purchased from their local Walmart.

There's a profound difference between your optimism, and Baron's optimism. Actually, to put it simply, Baron is overly optimistic, to the point where he is oblivious to the facts people present to him.

Be optimistic, don't be Baron-optimistic.

mmmmm.... brooooooood....

All well and good, but what about 45nm? I doubt AMD is going to wait until late 2009 to upgrade their process technology, as they've already said they're going 45nm in 2008.

How about a standardization of power benching?
http://www.dailytech.com/SPEC+Hope [...] le9988.htm

That should remove the pseudo aspect, as it already has in other types of benchmarking. Then any idiot can do it, which kind of removes the science aspect as well, but that's beside the point

Yet, AMD has not manufactured a working 45nm chip at this point.

http://www.xbitlabs.com/news/cpu/d [...] mpany.html

It takes about a year between tape out and mass manufacturing, if everything goes well. Considering AMD's background, I would expect them to debut their 45nm Shanghai in 1Q09, in limited quantities. Then it would take about one to two quarters for them to ramp the process node to mature status. It'll be 2H09 before we see those chip is volumes.

But of course, I'm just speculating here.

I believe that would be you who is tooting out is own arse.
Let's See..........
1) You try and use big words out of context.

2) The only claim you make is beyond the point of being silly that observing the make of the HSF is telling.
This tells you nothing. You don't know the makeup of the HSF, the speeds its designed to run, or the CPU operational range it's designed to handle.

3) You dismiss the claims of others which they have tried to establish using facts w/o actually countering any of their facts or providing facts of your own which are contrary.

All I can say if you are Mr Arse Speaker Himself.....

Sure. Take it out of context. It was a joke. Look at some of the names for Phenom and Barcelona.

I really wish you wouldn't third party me. I'm not even optimistic. I'm realistic. On the one hand you and TC say you hope it's faster, but you argued to hell and back when I showed where it was faster.

What was said was they haven't SHOWN any. It also says they are due to sample in Jan. But remember that Shanghai is just a shrink so a "tapeout" as such won't happen. Just as with Brisbane. They are having an analyst day Thursday, so maybe they'll shed some light on Shanghai.

Also, they have shown SRAM wafers earlier this year. Fabtech has reported that they are installing pilot lines in Fab38 and have some 45nm equipment up and running in Fab 36. But if you look at K8, they never missed a release date so why do you think they'll never reach such a milestone again?

You sure about that? Let's see what's realistic here. Barcelona is still at 2.0Ghz, with little to no availability from OEMs. AMD just ordered a "stop ship" not too long ago. On the other hand, Phenom is currently at 2.3Ghz, and suffering from TLB errata. To add oil to the fire, it'll be Feb before B3 stepping show up, with outrageously high TDP. Oh before I forget, AMD also revamped their roadmap to include K8 in their product lineups next year. Compared to Intel's offering, which has Xeon at 3.16Ghz, and Yorkfield at 3.0Ghz, and 45nm with Hi-Krap, do you really think AMD can compete in this state?

TC, ryman, and others, as well as me, HOPE K10 will be faster, because realistically, currently its not.

Do you really think Shanghai will be a direct die shrink from Barcelona?

From xbitlabs,

So nope, its an enhanced architecture. Kind of like Intel's Conroe => Penryn. Penryn was taped out exactly a year ago.
http://www.xbitlabs.com/news/cpu/d [...] 54338.html