IIRC, the IBM POWER chips shut off parts of the processor that weren't in use at a particular moment in time. I seem to remember seeing heat maps of the processor showing the effect on power consumption by sections of the CPU chip when that particular portion of the chip wasn't active. I don't know if IBM kept this up or if it got migrated to the Telum or Telum2 chips. I'm guessing it didn't or IBM would have been touting it.
Rex -----Original Message----- From: IBM Mainframe Discussion List <[email protected]> On Behalf Of Charles Mills Sent: Wednesday, April 16, 2025 3:07 PM To: [email protected] Subject: [EXTERNAL] Re: z17 On Wed, 16 Apr 2025 12:17:11 -0500, Paul Gilmartin <[email protected]> wrote: >the chassis gets hot and the fan turns on. Long ago I learned here >that IBM processors never reduce power. Is that still true? I have no idea but I would throw out there that most PCs are idle much of the time; most z processors are not. Everyone is now focused on "being green" so I suspect IBM may now find it important to reduce power consumption when they can, unlike in the old days, when the electric bill was a footnote to the cost of a datacenter. >Do the more complex processors still incur an energy cost for speculative >execution? I don't really know but it's hard to see how "doing something -- anything" would not require energy. CM ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
