Replies to post #61104 on Advanced Micro Devices Inc (AMD)

[chipguy]

Intel Montvale speeds reduced: http://67.19.9.2/?article=25591

So after Montecito wasn't going to get to 2.0GHz (1.8GHz at start) if this turns out to be true Montvale speed reduced from 3.0GHz to 2.1GHz???

Do you think this might actually be true?

I don't know where the 3 GHz figure comes from, the
highest clock speed for Montvale I had seen was an
implicit 2.6 GHz in an SGI deal. Whether that was a
holdover from a 65 nm Montvale I am not sure.

Right now Intel and IBM are playing a high stakes PR
game of blindman's bluff with respect to Monte/vale
and POWER5+. Both want the other to put their cards
on the table first in order to gain a tactical positioning
advantage.

The Montecito design can do 2.0 GHz (2.2 with Foxton)
at 100W TDP. How far Intel wants to take Montvale in
frequency depends on how much Intel is willing to let
the TDP figure rise and how much effort they put in
a tweak respin of Montecito. At the long held 130W
limit for the IPF family Montvale should be able to run
at 2.2 - 2.3 GHz (plus 10% Foxton) without a respin.
From Intel's perspective they would like IPF to beat
POWER5+ in performance without doing a full respin
for Montvale (to concentrate engineering resources on
getting the Tukwila to market as fast as possible) and
by raising its TDP above 100W as little as possible
to maximize its performance/power advantage over
POWER5+.