By the way...
> (It's depressing that these numbers have hardly moved since August ---
> at least on this test, the work that Robert's done has not made any
> difference.)
I merged some results and created another graph.
Please find an attached png file and you'll see obvious
performance improvement...
All the results are run on the same system:
Power 750 (32 physical cores, virtually 128 cores using SMT4)
mem: 256GB
OS: RHEL6.1 kernel 2.6.32-131.0.15.el6.ppc64
gcc version 4.4.5 20110214 (Red Hat 4.4.5-6)