I just don't think I'm reading this the same way they want to tell it. The 90nm process is fine, yet the FX-55 that was supposed to be based on it won't appear until 2005? The 130nm is "good enough"? I would think if it were fine, you'd be happy to introduce it on your top end chip, and you'd do it on the roadmap you originally promised in Q3 of 2004. Ok delays are always the way of things we understand that. But the idea of "introducing" a 90nm FX chip in 2005 right on the heels of a dual core chip smacks of one design being late just as the next design is coming on line. Using an advanced process in a lower end chip is typically a way of "working out the kinks" in the process before you move it to a more advanced chip. ATI uses this method. So if AMD is only introducing lesser chips on the 90nm process, does this really say "everything is fine"? That's not how I read it.

So exactly who is going to buy a 90nm chip that is 200mhz faster than last year's model in 2005 when the dual core will be just a few months later? No one would UNLESS the dual core was going to be late too...

So honestly the 90nm FX is going to be a year after its original planned arrival, and it seems reasonable to expect that the dual core FX will be the same. I don't think Intel is going to be sitting on its butt this entire time. (laughter) But with Intel's moves so far in 2004, maybe AMD isn't so "advanced" after all and would happily rest on its laurels. The AMD64 is a good series no doubt, but I am not believing the "everything is fine" they seem to be saying here.

Sorry, not buying this at all. There is no logical reason you wouldn't take the FX a top selling chip with superb margins and make it from a cheaper per chip process. You'd expand your earnings! UNLESS you were having issues producing the FX or your top end chips. Then, if you knew your yields were going to be rather poor, or maybe more to the point your yields at the top end speed, were going to be poor, you'd go for moving the lower end to 90nm instead. That way you'd have revenue to help pay for that 90nm equipment you bought while you tried to figure out how to fix the top end chips to run on the 90nm process.

Simple fact is, the Hammer itself was almost 2yrs late. The 2.4 FX was supposed to be here last year, not this, and the 90nm FX was 'supposed' to be here in the end of Q3 beginning of Q4. None of these have come true. The process is always late by at least a year, and I believe strongly you won't honestly see dual core AMD64 based cpu's until 2006 at the very earliest. These pipe dream press releases are just that.

It's a wonderful chip, but the future is going to take years to come into fruition, and by then the supposed savings you might have in any of today's technology, any "part at a time upgrade' that you thought might work will be long gone.

Notice the pattern. 90nm FX in late Q3. Ok 90nm FX in Oct. Ok, ok 90nm FX in 1st HALF of 2005. Meanwhile they sell you a chip built on a process that is 3yrs old plus. I don't think 90nm in the high end AMD64 chips is going to show till the END of 2005. And the LONG series of "don't worry you'll have it in Q X of Year XXXX" is the same old song and dance and I'm kind of tired of it.

/sigh