On Sun, 5 Nov 2006 16:31:06 -0000, Phil Payne <[EMAIL PROTECTED] RESEARCH.FREESERVE.CO.UK> wrote:
> >a) IBM didn't have a clue how much power an Amdahl (or any other pocessor) had. Amdahl (and I >was part of the process) "declared" its processors to IBM. Ah, yes. Thanks for reminding me. But IBM did have a clue, at least in later years. "Many IBM compatible processor models are also measured for the LSPR, whenever machine time can be obtained." See, for example: http://www-03.ibm.com/servers/eserver/zseries/lspr/os390t2.html > >b) You obviously know NOTHING about emulation, or at least commercial emulation. Thank you. >Fundamental's FLEX-ES is controlled these days by a USB dongle that defines the permitted >performance and number of engines. Yes, you can rehost onto a faster Intel processor or add >engines - but FSI's code works with the dongle to deliver exactly the licensed "MIPS" on the >number of engines licensed. You buy 8 MIPS (actually 7.9) and you get 8 MIPS. There is no >uncertainty in the process whatever. It sounds to me as if you're saying that MIPS is not meaningless on emulated processors, but very precise. Somehow it's difficult to believe that you'd be saying that. I can see though that the dongle could provide the capacity control or regulation that would be required. These days, with measured usage perhaps the dongle could be used to provide the capacity information needed to z/OS for the change in charging that would go with an upgrade to the underlying hardware. Tom Marchant ---------------------------------------------------------------------- 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
