Peter said, in part:

>I was hoping that the CPU speed would give a general idea of
>the performance - I know there's a lot more to it than that
>but I'd assume HP would match the CPU speed with the rest of
>the system.  Got to start somewhere.

Ok I'll bite.

Yes - on a relative basis between similar systems running the
same os you might be able to derive some reasonable assumptions.

>I did do a very simple benchmark program based on the kind of
>things our code does.  It's just that a customer wants their
>code running on MPE to run just as fast on UNIX. But I'd like
>to compare similar hardware.

The problem comes when you compare mpe vs unix if you are doing
any database work. If you are doing flat files with no database
then you might be able to get some comparisons. However, if
you are taking an app that uses TurboIMAGE to a RDBMS on the
*nix box then you are in for potential performance problems if
you are using identically performing cpu's. The overhead for
RDBMS access vs TurboIMAGE is going to consume lots of cpu and
memory. If the *nix box is faster and has more memory then you
might be ok, but if you took a TI app on a 120mhz mpe box and
ran it on a 120mhz hp-ux box with similar amounts of memory
you might find the hp-ux system slower. Depends on how the
database stuff is structured and what percentage of time the
app spends dealing with the db. You could mitigate that some
by adding more memory or by forgetting the older hp-ux box and
getting a newer/faster one since they don't cost that much anyway...

duane

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