Tench,
Yes.. this doesn't make sense..
Ya know last year this same division was saying something
similar. What I don't understand is why the PA-RISC side
of the house isn't on a shorter leash.
"Hold off until McKinley"
If Merced was decent, why the statement? PA-RISC would
be comparable.
If Merced sucks, why the statement? Stomping on IDF and
Merced news to come (real soon now) wouldn't be the
politcally correct thing to do.
Regarding the software base. I don't follow. How well
would Merced do re: native PA-RISC 8600 performance?
Probably not well at all. So who then "is the guinea
pig" to suggest "let's roll in Merced boxes to run
HP/UX" ... they would probably be laughed out of the
conference room. Let's roll in a Merced box to play with..
okay. But why not wait until McKinley as they would have
to do a recompile anyhow to see grand performance, right?
VLIW and all that, right?
Other issues.. how to establish a software base, etc.
Since HP/UX 11.0 supports clustering, high availabilty
servers, etc. etc. you can't roll-in and incorporate
a Merced based box into an existing cluster, different
binaries. You wonder how Merced HP/UX does at all.
What I am guessing is HP/UX binaries running on Merced
sucks as far as performance leaving HP/UX users in a
catch-22 regarding Merced outlined above.
Sound plausable?
Question comes back:
Who uses Merced and why? We've had this discussion
several times here.. Obviously not HP/UX.
Again, (re-asking our question):
What is Merced's target market and why?
Win64
Linux (this one is very questionable, IA32 better price,
Alpha much better performance and much
better price ;-)
????? Fill in the OS here (Monterey?)
Rob |
