>I did not mean to imply that it was -meaningful-. >Instead, in the proper context, such a number can be -useful-.
With all due respect, for what? If it's not meaningful how can it be useful. LSPR/zPCR have, unless it's change recently, only tested to 32 CPUs. The rest they 'straight line' the MP effect, which is bogus! The MP effect is a regression curve and NOT linear. I asked our IBM rep, a few years ago for LSPR for machines above 32, and he told me that you had to use zPCR for that. Since they don't test all workloads, and they project incorrectly after 32, what validity do these figures have? And, the 'Technology Dividend' just obsfucates the issue. If understanding the capacity of a given family/model is tough for a capacity professional, what happens when a senior manager has to make a decision to spend millions based on a meaningless indicator? I don't know about you, but it scares me, since it impacts the bottom line. I've had arguments since IBM introduced tier-based pricing in 1984, and I fear they will never go away. But, I think we have to convince IBM, and the ISVs, to go back to flat-rate pricing. Hah! NOT in my life-time! - Ted MacNEIL [email protected] ---------------------------------------------------------------------- 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
