Path: utzoo!utgpu!watmath!att!tut.cis.ohio-state.edu!bloom-beacon!mit-eddie!uw-beaver!blake!Tomobiki-Cho!mrc
From: mrc@Tomobiki-Cho.CAC.Washington.EDU (Mark Crispin)
Newsgroups: comp.arch
Subject: Re: DECsystem-20 (sort of...)
Message-ID: <3151@blake.acs.washington.edu>
Date: 8 Aug 89 19:12:50 GMT
References: <2679@phred.UUCP> <24933@tau.mips.COM>
Sender: news@blake.acs.washington.edu
Organization: Mendou Zaibatsu, Tomobiki-Cho, Butsumetsu-Shi
Lines: 43

In article <24933@tau.mips.COM> wilkes@mips.COM (John Wilkes) writes:
>I think
>the world's fastest PDP-10, the Foonly F-1, is still in service somewhere
>in southern California.  Supposedly, it was being used to do animation and
>graphics.  (Remember the TRON instruction?  Remember the movie?  Perhaps
>apocryphal, but the story is that the F-1 was used to make the movie.)

As I remember, the F-1 (only one of which was ever built) was at III (is
this right?) and was used to calculate the vector graphics used in TRON.
On a PDP-10, TRON (Test Right half of register with immediate, set masked
bits to Ones, and skip if any masked bits were Non-zero) is opcode 666!


>Tymshare, Inc., had a relationship with Foonly to make and market the
>"26KL" system, a Foonly F-4 that ran Tops-20.  The Foonly F-4 was based on
>2901 bit-slice processors, and it had a really cool console computer (based
>on a 6502, I think) that was capable of single-stepping the microcode.

The second fastest PDP-10, the Systems Concepts SC-30M, uses a 68K as its
console computer.  You could do all sorts of stuff from the 68K as well, but
it's been 5 years since I worked there and I forget the details.

I did the first port of TOPS-20 to the SC-30M including writing the FA10
(IBM disk channel/controller) driver and converting the USC SA10 driver to
use the IU10-based SA10.  The original SA10 (SC's disk and tape channel) had
its own IOT interface but on the SC-30M the SA went through the IU10 (which
was originally an IOT interface for the FA).

After all that, SC's customers told SC that they wanted to use their DEC
Massbus discs and SC ended up building a Massbus interface.  They also
abandoned their design for an Ethernet interface in favor of one that looked
like a DEC one.  This gave the advantage that unmodified DEC TOPS-20 kernel
binaries would run on the SC-30M.

SC designed an SC-40M on paper that (I think) would have been faster than an
SC-30M.  I don't think they ever built one.

Mark Crispin / 6158 Lariat Loop NE / Bainbridge Island, WA 98110-2020
mrc@CAC.Washington.EDU / MRC@WSMR-SIMTEL20.Army.Mil / (206) 842-2385
Atheist & Proud / 450cc Rebel pilot -- a step up from 250cc's!!!
tabesaserarenakerebanaranakattarashii...kisha no kisha ga kisha de kisha-shita
sumomo mo momo, momo mo momo, momo ni mo iroiro aru
uraniwa ni wa niwa, niwa ni wa niwa niwatori ga iru