Paul,
<Basically, the article is a HATCHET job - and completely out of date.>
Why do you say that? Is there something in the article
that shows that it is out of date?
<MERCED WORKS>
The author never said it didn't and said it was on track
for Y2K roll-out. His over-arching issue is performance.
The great unknown. As Linley Gwennap states in that article:
forbes.com
"There's been an acknowledgment that
Merced isn't going to be as powerful as
originally planned"
But we kinda knew that didn't we? I mean after all the
HP people have given us clues by comparing HP/UX binaries
running on Merced, 85% of 8500 numbers.
<It was late - so what - all of Merced's competitors ARE or WERE late>
But Paul, come on. Merced is so late it isn't even funny.
We see people dredging up articles from 1996 talking about
1998 Merced roll-out,etc. etc.
<HP is touting their upcoming NEW PA-RISC chips - now all they have to do is DESIGN THEM !>
Now you are really stretching. Why not find out what
you are talking about? The 8600 is just a clean-up and
shrink of the existing 8500. A done deal, nothing
dramatic at all. Like milking that P6 core, right?
hp.com
"The PA-8600, based on the award-winning PA-8500 core, will be used in HP's enterprise-class workstations and servers."
<That puts Merced in excellent shape.>
Why, who is going to use it and why? HP to run PA RISC
binaries? Certainly won't outperform the 8600.
Win64 on Merced? Do we REALLY believe THAT will be rolling
late next year?
Who is going to use Merced and why?
<My guess is that Merced will be a runaway hit>
Sure, but no one has figured out what will be ready to run
late next year or who is even going to be using it.
Certainly will be much less powerful than shipping 21264 and
considerably more expensive, less powerful than PA8600, Solaris and UltraSparc III... Solaris isn't even a blip on Intel today...
Who runs Merced and why?
Rob |
