Experimental Physics and
Industrial Control System

"Jeff Hill" <[email protected]> · Mon, 9 Aug 2004 11:08:23 -0600

John,

I have been quite busy the last few days, but have been intending to get you
some recent benchmark results the next time I have the distribution built
optimized here. At the moment I have the debugger set up to catch a bug that
takes a long time to occur and don't want to interfere with that.

You have to be careful with benchmarks as they can be easily skewed
depending on the situation. For example, when running client side benchmarks
on my workstation against my PPC based IOC, I suspect that you will mostly
see the performance of my IOC which isn't particular fast compared to other
systems. In particular, you probably won't see anything remotely related to
the maximum performance of the client library as the test will probably be
limited by a saturated CPU in the PPC IOC.

Jeff

> -----Original Message-----
> From: John Maclean [mailto:[email protected]]
> Sent: Monday, August 09, 2004 10:33 AM
> To: tech-talk
> Subject: PPC benchmarks
> 
> Does anyone have benchmarks for either a MVME5100 or MVME5500?
> 
> I'm interested in things like the number of records processed per second
> at n% CPU usage. Steve Hunt gave some figures recently for a number of
> systems but I don't think either of these two were included.
> 
> I'm also interested in CA benchmarks for PPC iocs. I came across some of
> Bobs benchmarks for 68k family processors but can't find anything more
> recent.
> 
> Thanks,
> 
> John.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System