[net.rumor] Foonly-10

mac@uvacs.UUCP (03/12/84)

A couple of years ago I heard about a machine called the Foonly-10, a
remake of the classic PDP-10 by an outfit called something like Foonly
Computers.  It made the news because DEC promised to sue.  Anyone know what
happened?  Did they win, or is the Foonly on the market?

Alex Colvin

 ARPA: mac.uvacs@csnet-relay CS: mac@virginia USE: ...uvacs!mac

joe@fluke.UUCP (Joe Kelsey) (03/16/84)

Foonly's come in several flavors, ranging in power from KA-10 to
(almost) KL-10 power.  I believe that Foonly was recently purchased by
Tymshare, Inc. and they are continuing work on a super-Foonly to
replace the now-defunct DEC-10/20 line.  Tymshare uses Foonly's for
most of their timesharing rentals and SRI also has a major inventment
in Foonly's.

/Joe

dave@uwvax.ARPA (03/21/84)

Well, in the ARPA hosttable, they list machine types, and FOONLY-10 is
listed for a number of computers, so I guess the beast is for real.

	Dave (I'm happy I use a VAX) Cohrs
	...!seismo!uwvax!dave
	dave@wisc-rsch.arpa

hans@log-hb.UUCP (Hans Albertsson) (03/27/84)

[]
There's supposedly a machine KL26 from Foonly, somewhere
between a 2040 and a 2060 in power, in less than the footprint
of 2020. This is supported by Tymeshare, using a TOPS-20
kind of OS.

Methinks I'd really like one of those, to replace this juvenile
delinquent of ours.

rpw3@fortune.UUCP (04/05/84)

#R:uvacs:-119000:fortune:9700003:000:3200
fortune!rpw3    Apr  5 02:33:00 1984

Ha! "Foonly" DOES come from (vaguely) the "foo(bar)" (fubar) word heritage!

I do not know the derivation of the name, but I do know about the machine
(a little), and it has a definite (if sometimes obscure) history.

DISCLAIMER: Most of what follows comes from the DECUS meeting at which
the first paper on the subject was given, plus a few hallway sessions
at DECUS several years later. Treat the political stuff as pure rumor.

Many moons ago, there was this computer architecture research project
at Stanford, mostly funded by ARPA but with some (?) DEC help thrown in
to build a much faster PDP-10 to run Tenex on. Now at the time, Tenex (an
operating system from BBN that ran on highly modified DEC KA-10's and later
to be ported/licensed/redesigned/"improved" into TOPS-20) was the darling
of the ARPAnet (much like Berkeley VAX/UNIX today). The DEC KI-10 was just
coming onto the scene, and was only 2x a KA-10 (or so), and the world ALWAYS
needs more CPU ticks! ;-}

Well, the KA-10 was built out of discrete PNP transistors, and the KI-10
was built out of (still fairly new at the time) TTL, and this Stanford
group set out to see how fast you could built a PDP-10 built out of ECL.
The name of the beast was the "Super FOONLY" project. (Ta Dah!)

They did a LOT of work, modelled and simulated and architected and designed
circuits and cache and E-boxes and M-boxes, etc., all with DEC looking over
their shoulders (not just at the design but also at the SUDS design automation
software), and were about to go build these things, and...

...ran out of funding. ARPA's priorities changed. Oops!

Well, here the story gets a little controversial, depending whose version
of history you listen to, and I am certainly no expert on it so I will just
say that not everybody was completely happy with what happened and how it
happened, but a little while later, the curtain rises again, and we see...

New processor announced by DEC, the KL-10, built out of ECL, with cache
and E-boxes and M-boxes, using the SUDS design package, and running about
five (5) times faster than the venerable KA-10! [Now to try and balance the
preceeding paragraph, the DEC guys say that the design was just a good paper
study when they picked it up and it had to have a lot of real-world engineering
put in, like making the ECL work at all, but like I said, I wasn't there.]

Anyway... back in sunny California, as usual there were people left over
who hadn't gone to work for DEC when they graduated who figured that much
of the work could be applied to a much smaller (but not much slower) PDP-10
architecture machine built out of 2901 TTL bit-slice ALU's.

Poof! Puff of Silicon Valley smoke, and Foonly & Co appears. Nice machine.
Real. Sold a bunch to timesharing companies who had been buying DEC gear.
I saw one once. TINY box (about PDP-11/44 sized). Neat. But also a small
company (with a funny name), so hard to compete with DEC outside the Valley.

I haven't heard from them in YEARS... (Then all this stuff pops up on the net)

Rob Warnock

UUCP:	{ihnp4,ucbvax!amd70,hpda,harpo,sri-unix,allegra}!fortune!rpw3
DDD:	(415)595-8444
USPS:	Fortune Systems Corp, 101 Twin Dolphin Drive, Redwood City, CA 94065