Newsgroups: comp.parallel
From: strauss@convex.convex.com (Henry Strauss)
Subject: KSR reliability (was Re: SMP vs. MPP)
Keywords: MPP, SMP, PVP
Organization: CONVEX Computer Corporation
Date: Fri, 6 Jan 1995 03:14:51 GMT
Message-ID: <3eb530$6bk@convex.convex.com>

In article <mjrD1n3BL.KuL@netcom.com> mjr@netcom.com (Mark Rosenbaum) writes:
>Much has been said of the business problems of KSR but I would
>be interested in hearing from KSR users and exemployees and how
>well the system was doing technically. My understanding is that
>there was a reliability problem that was either hardware or 
>software related.

Yes, there was a reliability problem -- in fact more than one --
but I don't see this as anything surprising.  The KSR machines were
still relatively new products, and AllCache the first hardware
implementation of virtually shared memory.

The KSR-1 was -- and still is (see also the posting of Melinda Shore) --
a pretty/relatively stable system.  The problems culminated when the
bad management practices became public and not enough people (inside and
outside the company) believed anymore that the company can make it.
At the same time, execs made it impossible to sell systems by issuing
stupid press releases.  In addition, Kendall Square tried to bring to
market the second generation (okay, let's say it was generation 1.5)
together with a new major os release which was not ready.  To summarize:
bad timing -- or: too many bad things happened at the same time.

In comparison to message-passing systems you also have to take into account
the much higher functionality of a general/multi-purpose parallel machine
like the KSR-[12] if you talk about reliability.

-Henry Strauss (ex-KSR analyst)
Senior Software Analyst
CONVEX Computer GmbH
Stefan-George-Ring 6
81929 Munich, Germany
voice +49 89 9302082
fax +49 89 935406
email strauss@convex.com


