|
This is exactly a problem...
Raw technical data is misleading, as I tried to show - like Mhz etc.
Another way is to measure system performance in relation to a well-defined
benchmark.
Unfortunately, benchmark results only show how well system performs this
particular benchmark, no more no less.
Benchmark results can only be extrapolated to a particular application, if
one understands very well, how *this* application relates to *that*
benchmark - which is very difficult and rarely the case.
Also, benchmark is taken in a particular configuration. Change
configuration of the same box - and suddenly results are different.
Too often benchmark results are taken as absolute measure - say, if system
A performs benchmark X better than system B, then it's faster overall.
However, with all the deficiencies of benchmark approach, it seems so far
the only "quasi-objective" way.
CPW numbers used by Rochseter are based on a strictly defined workload - in
other words they are benchmark-based.
(Don't forget that CPWs are measured for a balanced - most often this means
maximum - configuration.)
Unfortunately, there are too many different workloads nowadays. It's
impossible to aniticipate every use of a computer system and have a
benchmark for every case. Thus the attempt to alleviate this by publishing
Mhz ratings.
In this way one - caveat emptor - can estimate that if system has small CPW
but fast CPU, then *possibly* it *may* perform compute-intensive code
better than it's CPW rating suggests.
In short - it's so complicated...
Alexei Pytel
"Pessimist is nothing but a well-informed optimist"
"Steve Richter"
<srichter@AutoCoder To: <MIDRANGE-L@midrange.com>
.com> cc:
Sent by: Subject: Re: how is system
geared down?
owner-midrange-l@mi
drange.com
05/11/2001 01:14 PM
Please respond to
MIDRANGE-L
I apologize also Alexie. My commisars comment communicates my low opinion
of
ibm mgmt, whose actions I feel are threatening to make my 15 yrs experience
on this platform obsolete.
you are a great resource for this list alexie. I have learned more about
the
iseries from this thread due to yours and nathans participation. and I
thank you for it.
Now, could I ask if ibm publishes the cpu speed of its systems?
This could be a relevant question. Example: stocking a low cpw system with
a
lot of memory may allow that system to not page as much on its slow speed
io
channels, hence perform better than its cpw rating. Or a system that is
used primarily by pgmrs compiling pgms. A compile is a cpu and memory
intensive application. A system used in such a way as this may out perform
its cpw?
best regards,
Steve Richter
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2025 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.