The speed differences refer to the speed of the PB bus versus the PCI bus.
The individual controllers aren't any faster (e.g. 20mbs for fast/wide
SCSI), but they are less constrained by the bus they share.  Most larger
HP3000's have more than one PB bus, but it still is a bottleneck.  There
will be support for ultra SCSI sometime.

There were, in fact, comments about this improvement when the N-class
machines were first announced.  Postings from a couple of software vendors
(Robelle? Adager?) remarked about their early testing with this platform and
the remarkable improvement in performance.  I don't recall whether those
postings specifically stated that the PCI bus architecture was responsible,
but I believe there were references to "faster I/O".

-----Original Message-----
From: [log in to unmask]
[mailto:[log in to unmask]]
Sent: Wednesday, April 25, 2001 2:16 PM
To: [log in to unmask]
Subject: MPE 7.0 & hp-pb


I just received my entitlement announcement for the HP MPE/IX 7.0 bora
bora release.

I notice in the New Features and Technologies section that one is "IO
throughput performance improvement at 440MB/sec from 32/MB sec on the
current hp-pb (nio) systems"

Dang! That's a rather large increase and I am surprised it was not
remarked on in this group.  Does anyone know how this was accomplished or
was left so slow to begin with?

My 918LX always seemed grossly slow parsing through LIST'd LOGTOOL entries
with grep in posix, I wonder if this will help.

Paul

remove the _n.o.s.p.a.m
--

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *