On Thu, 5 Sep 2002 17:13:45 -0700, John Clogg <[log in to unmask]> wrote:

>>My point is that HP decided (for reasons that may or may not be valid) to
>sell a box at this performance point.  If they had accomplished that level
>of performance by building hardware that was actually that slow, no one
>would be complaining.  (They might not buy the machine, but they wouldn't
>complain about its existence.)  Knowing that the box is capable of better
>performance seems to be the issue that annoys everyone.  My opinion, on the
>other hand, is that if HP wants to sell a box at performance point X, and
>sell it for price Y, it really doesn't matter how they accomplish it,
>unless the choice they made caused the box to be more expensive.  I
>believe HP's choice to use software crippling actually kept costs down,
>because they didn't have to design multiple hardware platforms.

OK, for sake of argument let's say that the machine is really effectively
110MHz out of a possible 440MHz. (I wonder what measurement method HP used
vs. Gavin's estimate?)  Why is a 110MHz performance point so important when
current technology is in the 400 to 700 MHz range?  This means that the A-
class is really just a virtual 929, so why spend money to upgrade to it
instead of buying a used 9x9?  Heck, buy 2 or 3 used boxes for the price of
the new one.  No wonder HP thinks the customers have abandoned them for the
used market  --  HP has given them every incentive to do so.  After all, we
customers are in business too to make (save) money, not just HP.

As you point out, it makes no difference what method is used to provide
inferior performance, it just matters that they did.  In other words, the
technology provided is obsolete from day 1.  If they had come up with a
110MHz processor it would still be obsolete and the result would be the
same.  So the real problem is with the effective speed and not so much the
method of implementation.

I agree that the software solution is cheap to do, but it is also the most
irritating to the customers (does it matter what customers think?)  --
customers which HP is supposedly trying to help with a final performance
boost.  I think they are really worried that they might boost the customers
to far and delay prospects for migration to HP9000's.  This is a simple
stick and carrot approach to marketing strategy, and has nothing to do with
saving customers any money.

Give me a break HP.

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