On Fri, 6 Sep 2002 09:52:55 +0200, Christian Lheureux <[log in to unmask]>
wrote:

>To the best of my knowledge, software crippling has been used for years -
>about a decade. It was used to turn a 928 into a 918 (48 to 34-equivalent
>MHz), a 995/100 into a 991, a 979/100 into a 929/030, etc.

I am quite aware of the previous software crippling for HP3000 systems such
as the 918 to 928, etc.  However, here we have an entirely different
situation.  ALL current HPe3000 models are crippled!

In the past there was always a way to upgrade a lower level (crippled) CPU
to the full speed (uncrippled) CPU performance.  This, of course, allowed
them to "create" multiple performance points with fewer actual processor
models.  With the current HPe3000 offerings (A & N Classes) there is no
path, other than converting them into HP9000's, that will achieve full CPU
performance.  The fastest speed obtainable with the HPe3000 N-Class is an
effective 500MHz with PA-8700 processors (using HP's calculated
performance).  This stunted result is achieved with a CPU capable of at
least 650MHz.  No matter how much you are willing to pay, HP will not sell
you a full speed HPe3000.

To put it another way, the slowest low-end entry point PA-risc HP9000 you
can buy today is 650MHz, while the fastest and newest (just introduced with
great fanfare) HPe3000 is only an effective 500MHz using the same
hardware.  Meanwhile, the entry point low-end PA-8700 A-Class HPe3000 is a
mere effective 150MHz  --  thats 500MHz difference between HPe3000 and
HP9000 low-end entry points!

Why has it been concluded by HP that HPe3000 users have no need for more
speed?  What justification could there possibly be for this difference in
startegy?  I am sure this strange product line is not at the request of
users to keep their HPe3000's throttled so they don't get out of control.
Of course, MPE/iX uninhibited would run rings around a similar speed
HP9000  --  it would be most embarassing for HP to have to admit that the
obsolete MPE/iX is faster than the modern wonderful HP-UX!  I must conclude
that it has nothing to do with keeping prices low or some technical issue
or other altruistic claims.  This is simply a mis-guided marketing strategy
gone wrong with the goal of getting rid of MPE/iX in favor of HP-UX.

On top of all this, you get to pay more for the HPe3000 model than the
equivalent HP9000 model to which they will graciously demote your HPe3000
to at no cost, whenever you are ready.  As was pointed out previously, it
makes more sense to sell the HPe3000 machines used, as is, rather than
convert them to HP9000's.  You could probably buy a new HP9000 with the
proceeds and have change left over (depending on market timing).

So much for the final performance boost to carry HPe3000 users through the
migration/EOL/EOS period.

John Wolff

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *