AMD started trouncing Intel with the release of the Thunderbird back in 1999, 2000. Right up until the release of Intel's Core 2 Duo, they offered better gaming performance (a big deciding factor for many buyers) & better value for money. Intel's dual-core Pentium D range was a seriously under-performing, heat spewing disaster.

I go where the performance vs money ratio is right & currently own a Core 2 Duo. Prior to this I have been an avid AMD supporter.

AMD still offers excellent price-to-performance parts & Intel have yet to match their memory bandwidth.

It would be a fool that writes off AMD right now, as they were the underdogs for many years before they hit the limelight. Intel has surfed the wave of its shrewd marketing for many years, milking the consumer in the process.

I can't wait for AMD to make its resurgence & give Intel another run for its money, thereby increasing the competition in the market & increasing value for the consumer.

AMD doesn't only manufacture processors for the desktop & server markets either. Think before you leap.

If IBM buys AMD, Intel will really have some serious competition on its hands.

If you look at the author's word's, he clearly states this is an optimistic scenario, which, as you read further, you can see he himself may not believe as the most likely outcome. Simply, it is a projection "what if AMD strikes a jackpot, and Intel stands still" - both equally unlikely, but still a scenario that you have to be aware of coz, even then, Intel still seems to win hands down.

In "fact," I see that AMD will magically have increased from 2.3 to 3.2 ghz without changing TDP.

And to "power is what matters," 2.0 and 3.2 ghz would matter for quad cores since who buys quad cores to play cards!

Get a freakin dual or tri-core for that.

This article just described the absolute best possible outcome for AMD this year. But we all know that little of this will really ever happen.

There WILL be delays,
There WILL be bugs,
There WILL be disappointments,

but after all this, I hope that

There WILL still be AMD.

It's pretty obvious AMD has lost the initiative at the high end for now, maybe forever, or until Intel makes another Pentium IV-style goof.

A quad-core running at 3 Ghz iz enough to run anything.

I don't mind if AMD is 20% behind at the top end, as long as they deliver great value at the price/performance sweet spot.

Yes, the article mentions Bloomfield, what is a high-end server variant of Nehalem you will perhaps see this year. Lesser variants will come only next year.

Last time I checked the 45nm process has been perfected, its just that Intel has more fabs on which to push chips out. 

But aren't it good to see we are back to Ghz willy waving. :( 

I'd rather have a chip that runs slower , but does more per clock, at the moment, we seem to be back to the silly old days lets not make it effiecent, lets just bump up the speed!.

After reading your analysis I think the title of the article should read "AMD makes a desperate 45nm stab at Intel's Penryn".

Wouldn't you agree?

Anyhow, the title did get me to click the page.

Call me a spoiler here, but I think you missed one variable in your calculations. 

Nehalem.

You did say September.

SPARKS

Nice article and interesting read nice to hear something from the AMD side.

Intel's already gotten some of their processors up to 5ghz on water, so i would be surprised, even if they stay with what they have now, if they got those processors well above 3.6Ghz.

This article basically took a rumor and stretched it to a scenario where AMD will do everything right in the coming year and Intel will do everything wrong....it's funny.

You even think to believe when AMD can hardly make larger gate sizes work that they are now going to be able to magically leap to 45nm? C'mon, get serious! This is nothin gmore than an attempt to grab ink (cyber or otherwise) to hope to delay procspective buyers from buying Intel now. An there is no way they can match the 45nm performance of Intel, because in this case it's not just size that matters unless they an IBM have teamed up to find a way to match the hafnium aided reduced leakage, right?

Ummm, Maybe you didn't see that:

http://hkepc.com/?id=757&fs=c1n

AMD 45nm Quad Core in Q4.

So I really don't what you're talking about with " quick 45 nm delivery ". It's ain't going to happen.

Time to stop dreaming.

This artical forgets that power useage is what matters. Most people can't tell the the different between 2.0Ghz and 3.2Ghz when openning word or playing cards online but if I could get another 15 minutes out of my battery when triing to finish a report or watching a movie then I would be happy