Describing memory as the new disk is quite apt. A given instruction may take between zero and hundreds of thousands of clocks. Z10 and later processors execute upto 11 instructions in parallel, although they rarely exceed three.
The key bottle neck is operand access times. L1 cache costs 1 clock cycle. L2 costs 7 clocks. L3 is 60 clocks and L4 is 600. A simple main memory access (assuming a TLB hit otherwise there may be 2 further memory accesses to look up the page table) will take 2000 clocks. If the operand is in a CF storage structure which is part of a GDPS Plex then access may take hundreds of thousands of clocks. If the processor cannot execute any instructions, it's in a pipeline stall. This is very common, with many commercial workloads experiencing stalls more than 60% of the time. Just think about that ... the wait bit in the PSW is zero, so MSU's are accumulating, but the processor is only emitting instructions 40% of the time. IBM's solution was to introduce hyperthreading. In z/Linux this gives a 50-80% performance boost. They disabled support in z/OS reportedly because they couldn't agree how to bill it. No free MSU's for you :-) Matthew On Thu., 7 Mar. 2019, 15:19 Charles Mills, <charl...@mcn.org> wrote: > The 360/40 had one: 7 us for LR, 10 us for L. > > It is not possible now. A single instruction may literally add no time at > all to some instruction sequence. > > My imperfect model is that main storage is the new disk. Figure that > instructions take no time at all and memory accesses take forever. > > Charles > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On > Behalf Of Joseph Reichman > Sent: Wednesday, March 6, 2019 5:59 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: instruction clock speed > > Hi > > > > > > From what I remember there used to be a list of instruction clock speeds in > the principle of operation I wonder if that's still available > > > > thanks > > > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
