Oppolzer - Informatik / Blog


Blog-Hauptseite      Neuester Artikel      Älterer Artikel      Neuerer Artikel      Älterer gleiche Kategorie      Neuerer gleiche Kategorie

ASSEMBLER-L - Optimierung von ASSEMBLER-Programmen

Subject:

Re: Looking for Diagnostic Process for Strange 0C1

From:

Bernd Oppolzer <bernd.oppolzer@T-ONLINE.DE>

Reply-To:

IBM Mainframe Assembler List <ASSEMBLER-LIST@LISTSERV.UGA.EDU>

Date:

2010.06.05 01:30:27


I recently removed a unneeded loop in a database interface module from a
software vendor which was executed on every database fetch. In some IMS regions,
up to 5 per cent of the total CPU was consumed in this loop. The elimination of
this loop (after at least 5 "active" years) will save us CPU costs of 300.000
euros per year.

Kind regards

Bernd




-----Original Message-----
>> From: IBM Mainframe Assembler List
>> Sent: Thursday, June 03, 2010 4:19 PM
>> To: ASSEMBLER-LIST@LISTSERV.UGA.EDU
>> Subject: Re: Looking for Diagnostic Process for Strange 0C1
>>
>> These sort of discussions have been confusing me for some time now.
>>
>> Why is performance to this extent so important?  Take me for
>> example.  I love writing in Rexx.  It is a nice language, and
>> it is fast.  If it isn't fast enough then I look for ways to
>> improve it, otherwise it is all fine.
>>
>> There seems to me to be a lot of talk and effort on this list
>> on how to shave off a millisecond here and there.
>>
>> Your worst assembler program will most likely out-perform my
>> best Rexx exec any day.
>>
>> Why does it all matter?  Ok, if I am writing a compiler, then
>> that is a different story.  But how many people who talk
>> about pipelines and this instruction and that instruction
>> need that level of performance?
>>
>> I'm not being facetious. I genuinely am curious to know.
>>
>> Kindest regards ...
>>
>
>

Blog-Hauptseite      Neuester Artikel      Älterer Artikel      Neuerer Artikel      Älterer gleiche Kategorie      Neuerer gleiche Kategorie