On Tue, Nov 17, 2009 at 12:56 PM, Tony Harminc <[email protected]> wrote:
> The "specialty engine" part is just so much marketing hype. As has > been discussed endlessly here, the current specialty engines are the > very same engines as all the others, but with different legal Ts & Cs > that make it impossible to run "traditional" workloads on them. IBM > marketing tries very hard, without actually saying so, to suggest that > these engines are different hardware that somehow make Java,Websphere, > Linux, DB2, or whatever run faster. It may well be that specialty > engines can save money under some conditions, and there's nothing > wrong with that, but it's annoying to hear so much borderline stuff > about them. > > Of course IBM *could* put truly different engines in if they wanted > to, and there have been rumours for a long time about putting CBE > chips and such into System z boxes, but I don't believe it's happened > yet. > Don't forget that the zAAP/zIIP/IFL microcode disables ONE diagnostic instruction that z/OS uses during initialization, so it's (slightly) more than T&Cs. I see this as a safety, so nobody can say "Gee, we didn't realize", as they'd have to hack the z/OS IPL process to get it to work. Which (as Harry notes) doesn't change your valid point. As for CBE, IBM was talking up Haplon and CBE on z a while ago, but that seems to have died down. Interesting. ---------------------------------------------------------------------- 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
