[Simh] Simh execution speed improvement - about 45% improvement.

[Zane H. Healy]

> I am serving 31 disks of 4GB size, and one disk of 40GB size. All those
> disks are hosted on a single 250GB SATA disk that is in one of
> the Two host desktops.  The SATA disks gives me a lot more performance than
> the Original SCSI disks connected to the vaxes.

Okay that means SIMH can handle serving up the disks that are currently
hanging off of my XP1000.  I wasn't sure how large of disks it can currently
handle.  An interesting adaption of your configuration would be at least one
more physical disk, and then use Volume Shadowing with the different disk
images in the shadow set living on different physical disks.  I'm currently
using 18Gb, 36Gb, and 50Gb drives on my system, but in a configuration such
as this I'd prefer stepping back down to 4Gb drives, as that would allow
them to be backed up periodically to cheap DVD-R blanks.
  
I'll have to find time one of these days to setup a test cluster, and see
how well this works for me.

> I plan to replace 4 vax 6000-640 each with 512MB RAM with those
> simulators. This actual setup is an experimental one, and we will purchase 2
> servers (probably DL380 G5 with quad core xeons) for the "production"
> setup.

I know someone using DL380's for PDP-11 replacement systems.  Good hardware,
we've been using the G3's through G5's at work, the DL360's are nice as
well.

> Thanks for the pointer to the intel linux compilers, you can count on me
> to try it, and inform people on the list about my results.

I look forwards to hearing how that goes.

Zane