-----Original Message----- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Ken Porowski Sent: Friday, July 13, 2007 2:58 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: Links to decent 'why the mainframe thrives' article
The processor clock speed may be slower but it is running a different instruction set and has the availability of other processors to offload some of the functions (Crypto, I/O, etc.). I would guess that a small stand alone 'benchmark' like you suggest would almost always run faster on the alternative platforms. What is needed is to fully tax the platforms capabilities using a 'real world' workload. But of course my world is different than yours so my benchmark is of little use to you as is just counting I/Os per sec, connections per sec, pages served per sec etc. <snip> I agree. This was once done with Univac to show that it was faster than the S/370. Mind you, when a floating-point intensive job was run on the Univac, it stole the show. But for general computing purposes where multiple tasks were done at the same time, the S/370 was the winner. So the point is, use a data base with some large number of entries, run the quarter-end processing for a "medium" sized construction supply house or auto-parts company, while doing all the other stuff that the server does. There has to be some benchmarking workload that could be put together to do throughput testing of a system to compare one to the other. I'm willing to work on the idea if someone has something on z/OS that could be ported to Windows. I don't have the COBOL compiler for *NIX environments, so I can't handle that one. Regards, Steve Thompson ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html