Mike,

Re:** Pricing Added **

Yay! thx <eom>

comb,
SocketA has 453 pins, a number that looks suspiciously like the 563. Ragardless, it looks more like a SocketA derivative rather than a Socket753 derivative.


isn't socket A 462 pins?

Do not listen to this guy, lost all credibility, he loves Intel just as much as he criticize AMD. His sources come from a "Russian" AMD engineer.

Mike,

Its very very fine to have your updates on inventory, now it would be even better to have the pricing with it. If its too much work, only give pricing for key parts, especially low-end AthlonXP

Anyway, thanks

One thing that everyone seems to forget : Intel can't just sulk Dell because of an infidelity, I mean Intel needs Dell's volume just as much as Dell needs Intel special pricing. Intel has special relationship with HP, Dell has no exclusivity so why would Dell be so faithful to the sultan and its harem.

Dell would be happy to stay with Intel as long as they don't lose customers, but now they are losing to Sun and HP and IBM and lots of whiteBox servers farms

edit: my take is that they waited to see Nocona and they are not very happy with it, and they have lots of benchmarks to prove it.

Bash where it hurts...

When coppermine was introduced, Intel wouldn't ship it at 550MHz .25u speed grade but rather 667, 733MHz. And, guess what, the pipeline wasn't stretched 50%.

First Northwoods would follow Presscott pattern but was never advertised at a frequency over Williamette's 2GHz. Right after the secod stepping was ready Intel announced and delivered 2.4GHz immediately, not the case with presscott.

While there may be some bashing, you're clearly on denial.

The stages after decoding didn't change at all.

I know that this is, roughly true. I just presume there was a timing stress at that level.

I think 90nm K8 is good for 3.2 GHz and I have always stated 130nm K8 was good for 2.6 GHz on Aces.

I have always beleived SOI was to bring at least some improvements over bulk 130nm, the two extra pipeline stages probaly relaxed stress on critical steps of the K7 core, and I was too shooting for 2.6GHz on 130nm SOI and 3GHz+ on 90nm at ace's, back in 2002, when everyone was worried by 800MHz hammer demo.

Now my new prediction is over dual-core frequency, which could be real nice, I'd shoot for 2x 2.6GHz+ in Q2'06 or before, especially for Athlon FX chips which will have more TDP headroom at 104W.

Keith,

I can't beleive it, not a single comment on improving ASP as the transition to Athlon 64 occurs, and Opteron volumes increase.

Neither any comments on improving flash margins due to the switch to 110nm and MirrorBit.

Given the flash memory and
manufacturing issues, we think investors would do best to
stay on the sidelines until AMD finishes the next two
quarters. Our recommendation remains Neutral.

He couldn't be less positive

Hi Greg,

No, no, after you. I insist. After all, you have been so very kind and respectful

If you insist,

I'd have a Jack on the rocks please, what a nice fireplace!

That depends on how you define profits

Simple, money invested vs money earned. No twisted reasonning there. Itanium have to generate billions per year to aceive that, and it's millions chips aways from that.

wbmw,

Alpha didn't survive because DEC and Compaq didn't manage it very well and HP didn't want to be a CPU design house. How is that similar to Intel's position with IPF?


My argument is that would Intel have depended on Itanium as much as DEC depended on Alpha, Both Intel and Itanium would have tresspassed already. Don't pretend Intel have managed it supperiorly, it's on design tables since 1994, released since 4 years and barely been sold a few 100's of thousands times.

Itatium is still far, far away from profits.

wbmw,

"I've heard this argument before, but time and again, the market proves that performance and positioning will rule over price/performance. That's why AMD isn't the size of Intel today."

Intel's size has more to do with some CPU choice by IBM, same story on the O/S side with MSFT.

AMD's ambition is recent. It begins with the K5.

I would say the Alpha argument is excellent, and wouldn't Intel be so rich, the Itanium project would have bankrupted them. DEC hadn't a cash cow line of commodity CPU to cover Alpha spendings.

Rupert,

Wonder why ML haven't wasted any time takling about the Server market.

Intel price target at $40, what is uncompetitive-AMD price target at ML? $18 or so

ML is filled with sh... erh.. Intel shares

Max