On Mon, Jul 20, 2009 at 9:51 AM, Bruce McKnight <[email protected]>wrote:
> I'm familiar with the discrepancy between hardware and software MSUs. IBM > has put the customers in an unfortunate position of contention with > software > vendors. > > I'm still curious about the gap between the 117 hardware MSU and the 124 > MSU I measured. Is it reasonable to attribute it to PR/SM overhead like we > do > for MIPS? > > Also, has anyone ever successfully defended that hardware MSUs are for > hardware sizing only and that IBM's software MSU rating is actually an > appropriate measure for ISV software? > > Reaching for a literary metaphor... "A word means (just) what I say it means; no more, no less: The Red Queen" and thus it is with MSUs. Put slightly differently, if you're looking for an exact tie-off between measurement data from your system and MSUs or MIPS, be prepared for a trip through the looking glass. -- This email might be from the artist formerly known as CC (or not) You be the judge. ---------------------------------------------------------------------- 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
