website/content/museum/scientific-computer-symbolics-3600.md

title, categories, types, origins, years, cpus, params

title	categories	types	origins	years	cpus	params
SCIENTIFIC COMPUTER SYMBOLICS 3600	Computers	Professional Computer	U.S.A.	1983	Custom chip	name company company_name specs comments oc_id source SYMBOLICS 3600 /companies/scientific-computer SCIENTIFIC COMPUTER label value NAME SYMBOLICS 3600 label value MANUFACTURER Scientific Computer label value TYPE Professional Computer label value ORIGIN U.S.A. label value YEAR 1983 label value KEYBOARD Full-stroke keyboard label value CPU Custom chip + Motorola MC 68000 label value SPEED 8 MHz label value CO-PROCESSOR Unknown label value RAM 2.3 MB label value ROM Unknown label value TEXT MODES Unknown label value GRAPHIC MODES 1100 x 800 / 1280 x 1024 label value COLORS Monochrome display ? label value SOUND Unknown label value I/O PORTS Centronics, RS232c (3), Ethernet label value BUILT IN MEDIA Hard Disk (170 MB) label value OS Unknown label value POWER SUPPLY Power supply built-in the system label value PRICE $110 000 (USA, 1983) author date text Allan Dobbins (United States) Sunday 27th August 2017 I used Symbolics LISPMs in the latter half of 80s to early 90s. They were far ahead of anything else. Many programmers felt that they were about an order of magnitude more productive on these compared to the alternatives at the time. The geniuses who designed the LISPMs (originally a prototype in the MIT AI lab) created a completely integrated environment in which the same keyboard shortcuts were available everywhere. The compiler was amazing: one could do incremental compiles e.g. just recompile the function you had just added or changed without recompiling the whole, possibly very large program. Of course you had to do a global recompile from time to time. We had two head systems $ color graphics on one monitor, and the Zmacs editor, etc on the other monitor. I could change my program easily without reloading, linking, etc. by using the incremental compilation capability. I''ve never experienced anything close since. Thirty years ago! Some may find this hard to believe, but others who used these systems can verify. In the mid-90s we created a weak imitation on Silicon Graphics computers $ a large C++ program with OpenGL and OpenInventor and the Guile scripting language on top that allowed one to do a lot of things dynamically with the underlying 3D graphics. Not as flexible though. I maintain that in some respects the Symbolics LISPMS represent a pinnacle of computer engineering that has never been equaled. author date text Christian Berger Wednesday 5th September 2012 Actually in 2002 the CCC had a talk on Lisp machines.\ http://media.ccc.de/browse/congress/2002/19C3-488-symbolics-lisp-machines.html Apparently on those the RAM was just cache for the disk. They didn''t have a file system as such, but used the LISP memory management for that. Apparently the web server of the white house ran on a LISP machine during the Clinton era. author date text GPUDude Saturday 30th July 2011 Yes, Symbolics later created NUBUS add in cards that worked with 68K macintosh Quadras called the Macivory, Macivory II, and Macivory III. If you want to pick one up for yourself you can email sales@symbolics-dks.com. Note that they''re pretty pricey. Here''s a screenshot of a macivory: http://www.sts.tu-harburg.de/~r.f.moeller/symbolics-info/development-environment/index.html author date text Ilgaz Sunday 16th January 2011 I saw one of them in very high end graphics/animation design company. It was happily showing 1080P uncompressed 3D animations in 1993 or so. Trust me, consider yourself lucky if you haven''t seen one as it would really shake your impression of later "breakthrough" systems. :) author date text Jonathan Wednesday 6th October 2010 The Symbolics machines are, in my opinion, the most ambitious computing platforms of their time so far. If you would like to see what they were capable of, please go to you tube and search "mind''s eye breaking the ice" (without the quotation marks). author date text kwa (us) Monday 28th March 2005 Didn't symbolics later build this onto a NuBus card that fit into a Mac II? author date text Qbasic DUDE (US) Sunday 2nd May 2004 Wow! This thing has amazing graphics even for 1990\ and it was made in '83!\ I wish there were screenshots though 230 url date https://web.archive.org/web/20230605030207/http://www.old-computers.com/museum/computer.asp?c=230 2023-06-05

This scientific computer had characteristics which were incredible. It had a 170 MB hard disk, when other computers used 10 MB hard disks!

It was designed to make CAD or artificial intelligence applications. It used a graphic interface very similar to the Macintosh one (!) and was sold with a mouse.

Several languages were supplied with the computer: Lisp, C, Fortran and, InterLISP (developed by Xerox).

It used a custom CPU, the 68000 was used to run the keyboard, the mouse and to boot the main processor. The main processor had an architecture derived from the MIT CADR Lisp machines.

In 1983, it cost about $110,000!

---

More information from Peter Grace:
I worked for Symbolics in the UK until the company went into chapter 11. I then bought the UK stock to provide support to UK and European users.
The 68000 processor was used to bootstrap the LISP processor.
The processor on the 36XX systems had a 36 bit wide word, 32 bits data 4 bits data type.
The processor was integrated onto a single CPU, the Ivory, then used 32 bit data , 8 bit data type. This CPU used in systems MacIvory, XL1200 and NXP1000.

Michael Haynie adds:
The Symbolics family of computers were designed to run LISP more or less directly. So the OS was also written in LISP. It was known as Genera. I still have the manuals for Genera 8.
The primary display is a monochrome bitmap display, black text on a green-grey-white background. Very easy on the eyes. Manufactured by Symbolics, Inc.
I also have a pair -- a 3650 and 3620, both later models.

Brad Miller specifies:
The native LISP that Symbolics ran was Zeta-Lisp and the built in editor was Z-MACS, a version of the emacs editor.
I have a pair of 3620 machines complete with monitors, keyboards and documentation that I bought at an auction in the mid 1990s. The machine used additional keyboard modifiers for extended commands; so the keyboards have super, hyper and meta keys in addition to the normal key set.
The earliest 3600s were about the size of an 18 cubic foot refrigerator. They stood about five feet tall by about 2.5 feet wide and three feet deep. The 3620s that I have are only slightly larger than current desk side Intel machines, but the Symbolics are much heavier.