[email protected] (Dana Mitchell) writes: > Well, yes I can, I can run it at home too.... > > My point was, when will it be uneconomical enough for IBM to quit > producing 'real' z CPs and only offer z machines running on power > cores?
re: http://www.garlic.com/~lynn/2014h.html#9 Demonstrating Moore's law http://www.garlic.com/~lynn/2014h.html#10 Demonstrating Moore's law http://www.garlic.com/~lynn/2014h.html#11 Demonstrating Moore's law http://www.garlic.com/~lynn/2014h.html#12 Demonstrating Moore's law http://www.garlic.com/~lynn/2014h.html#13 Demonstrating Moore's law for all we know, they may be close to that now. starting about two decades ago, i86 started moving to risc cores with hardware layer translating i86 instructions into risc microops ... as a result, they have now largely negated the throughput difference between i86 and risc processors. for a little other drift ... http://www.cringely.com/2014/06/04/decline-fall-ibm/ "The Decline and Fall of IBM: End of an American Icon?" loc1630-34: AFTERWORD What if Ginni Doesn’t Listen? Here’s what the IBM insider I quoted in my introduction says is coming today from Ginni Rometty’s office: “Ginni is betting the farm on PureSystems. She is also betting the farm on Cloud. The problem is she is blaming flagging hardware sales on Cloud-ification. ... snip ... a couple old posts "From The Annals of Release No Software Before Its Time" refers to zVM cluster when HONE vm370 had large cluster single-system-image in late 70s and purescale commerical/rdbms scaleup when we were to do ha/cmp cluster scaleup in 1992 http://www.garlic.com/~lynn/2009p.html#43 .. http://www.garlic.com/~lynn/2009p.html#46 .. i.e. 17yrs later ... aka ha/cmp 128-way cluster scaleup by ye92 supporting commercial, RDBMS, numeric/scientic, national labs, massive filesystem, etc; meeting on rdbms part in ellison's conference room jan1992 http://www.garlic.com/~lynn/95.html#13 other ha/cmp posts http://www.garlic.com/~lynn/subtopic.html#hacmp cluster scaleup email from the period http://www.garlic.com/~lynn/lhwemail.html#medusa by feb1992, the scaleup had been transferred, we were told we couldn't work on anything with more than four processors, and announced as supercomputer for technical and scientific *ONLY* in early 90s, we were working with non-ibm rdbms vendors for ha/cmp cluster because 1) ibm only had mainframe rdbms and 2) other vendors had common source base for their open systems and their vax/cluster implementation (i could simplify operation in ha/cmp cluster scaleup by providing cluster API semantics that resembled vax/cluster ... also a couple of these vendors had list of ten things dec had done wrong in vax/cluster that they wanted me to avoid). Part of the motivation for the transfer and announce for technical and scientific only ... could have been mainframe RDBMS (DB2) people complaining that if i was allowed to go ahead, it would be way ahead of anything they had -- virtualization experience starting Jan1968, online at home since Mar1970 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
