Cecile Chi said...


>The salesman predicted a possible 50% improvement, but warned
>them not to expect more than 30%.  After the upgrade, batch jobs
>ran in one-third to one-fifth of the time, and problems due to jobs
>interfering with each other or night-time processing not finishing
>before morning disappeared.  The users see an even greater
>improvement, because they measure response time from the time
>they submit a request to the time they get the result, without
>realizing
>that their job was in WAIT state for two hours before the upgrade.

Excuse me, HP salesman, have you looked at a performance chart
recently????

The one I have in front of me doesn't have the 937, (it has the 918
series), but a 928 is a 1.4 (relative to a 918) and a 969/100 rates 4.0.

If my memory serves me correctly (and it is only single bit error
correcting memory), the 937 would fall between the 1.0 and 1.4.

Cecile didn't mention what model 969 they bought, but a 969/100 is the
worse case.  You go up one notch to a 969/200 and the relative
performance is 7.1.

Now of course, YMMV so I can understand not wanting to overstate the
performance gains, but predicting 50% and expecting no more than 30%
improvment doesn't make sense to me.


>Let's hope the sales force doesn't counteract the effect of this
>advertising campaign.

Well the good news is that there isn't that much of a sales force out
there!


Way to go HP on the marketing push.  It was also good to see that the
CSY person (forgot her name allready) passed the message to Jeff V. to
forward to hp3000-l.  I guess we *are* making a difference.

/alan
[log in to unmask]

Alan "Been through lots of upgrades over 20 years.  Now where is that
Series II" Ambers