Not too shabby for $62K in 1952, this thing operates at .12Mhz has roughly 2K of memory and each tape holds around 360K.
Plus for all you case modders, it already comes with 200 glowing tubes. Try to beat that with your little LEDs.

to meet all your

Price $62,500
complete with tape drive and typewriter
Available 120 days*


MEMORY—magnetic drum, 512 word capacity. WORD LENGTH—30 binary digits and sign.

ARITHMETIC OPERATIONS—Addition; Subtraction; Multiplication (with round-off); multiplication (complete product); Division (with round-off); division (with remainder).

LOGICAL OPERATIONS— extraction; shift right; shift left; tally; overflow branch; conditional transfer of control (branch); halt; input and output operations.

MAGNETIC TAPE UNIT —single unit, using 1/2″ plastic tape on 1200 ft. reels. Data recorded in four channels including sprocket channel; 64 words per block, 1500 blocks (96,000 words) per reel. Programmed operations: read one block; record one block; move tape backward one block without reading or recording.

TYPEWRITER—Flexowriter fully controlled by machine. Programmed output operation calls for typing out blocks of 64 words on eight lines with automatic carriage returns and spaces between words, using octal representation, or typing out any number of words using decimal and alphabetic representation. Manually initiated input of single word to any desired address, or sequences of words with any desired starting address, Input may be either octal or decimal and alphabetic.

SPEED OF OPERATION— basic pulse rate 120 kcs., average time for internal operations 40 milliseconds, tape instructions approximately 1-1/4 seconds per block, typing out operation at rate of 10 characters per second.

EQUIPMENT SPECIFICATIONS—approximately 200 tubes and 2000 crystal diodes; power consumption approximately 3 kva., 120 volts AC. Main computer 6 ft. high, 16 sq. ft. floor area, mounted on casters. Control desk (34″ x 60″) holds typewriter, tape drive and control panel.

EASE OF MAINTENANCE – Construction is chiefly of flat chasses mounted on racks freely accessible on both sides making parts conveniently accessible for maintenance. Extensive use of standardized plug-in components permits rapid replacement and test and repair if needed of suspected components without shutdown of machine. Various manual controls are provided for the convenience of maintenance personnel including operation for one cycle or one instruction at a time, repeat of one operation, synchronization of test oscilloscopes, etc. Tape operation is checked continually by use of an auxiliary channel using the so-called “odd” pulse check per character.

Inquiries should be addressed to the Development Department

Founded 1949

Executive Offices:
160 Avenue of the Americas
New York 13, N. Y.

Laboratory and Plant:
265 Butler Street
Brooklyn 17 N. Y.

  1. Steve Rush says: April 13, 20083:01 pm

    Pretty amazing, for 1952. That “approximately 200 tubes and 2000 crystal diodes” gives a clue to its logic, and “30 binary digits and sign” implies sign-and-magnitude arithmetic. Had twos’ complement been invented yet? Of course, $62.5K was a fortune in those days, but IBM wouldn’t _sell_ you a computer at any price; you had to lease.

  2. Uncle B says: June 20, 20083:35 pm

    Most of todays computers are not used to solve engineering or social problems. Most computers today are used by prepubescents to chatter endlessly and aimlessly and by males to watch porn! We are certainly no further ahead, and arguably have gone backwards!

  3. Paul says: February 1, 20093:30 pm

    That’s a fair point. I sometimes wonder whether taking computers out of the laboratory and making them available to the masses has retarded our progress as a society. Back in the 60’s, it was anticipated that in the future, children would be taught computer languages from an early age. Of course, the availability of personal, easy-to-use machines which keep all the complex stuff from you means it simply isn’t necessary for most people.

  4. David Lubkin says: May 12, 20128:49 pm

    Elecom was the first mini-computer company. Its computers were designed by its president, Dr. Samuel Lubkin, my grandfather. He was an unsung pioneer of computing, whose substantial career in the field began when he was the engineer in charge of the ENIAC once Mauchly and Eckert delivered it to the Army in 1946, and continued with EDVAC, SEAC, REEVAC, etc. I inherited his papers, and will be making the interesting bits available on the web.

  5. Toronto says: May 13, 20126:44 pm

    Very cool, David. Is this the same Elecom that was bought by Underwood?

  6. David Lubkin says: May 15, 201211:52 am

    Yes. Sam was a brilliant engineer but a lousy businessman. (My family has a history of making *other people* rich.) Elecom failed because it was too successful — it was undercapitalized, and they didn’t have enough cash on hand for parts and people to fill their orders. After Underwood bought the Electronic Computer Corporation, two months after the ad above appeared, he stayed on to run the division for a few years. His patents at Elecom and Underwood wound up being sold to Curtiss-Wright, which licensed them to IBM, GE, Honeywell, RCA, Sperry-Rand, etc.

  7. Toronto says: May 15, 20125:52 pm

    David – I googled more about Elecom the other night and found a PDF of the instruction set manual online. I think I stayed up past one AM reading it! It reminded me of programming an NCR that came around about 10 years later (1962 – but I programmed it in the mid 1970s.) It had core memory instead of drum, but only about 1/4 as much as the one here. Both machines would actually generate self-modifying code as a matter of course! (eg, the “TALLY” or “for-loop” countdown instruction.) Hand assembly, barely above bearskins and flint knives, but dammit, it was fun.

    So, the Elecom-style methods lived on for a surprisingly long time in such a volatile technology.

    Thanks for the trip, and cheers to the memory of your grandfather.

  8. David Lubkin says: May 16, 20122:11 pm

    My grandfather died before I had a chance to discuss computers with him, so I know my father’s stories best, and my father embellished. I’m now going through his invention disclosures, patents, etc., to separate embellishment from reality. But I was told that Sam invented self-modifying code.

    Another invention I was told was his is clocked logic for computers. Designs had been asynchronous up until then — they’re inherently faster. But he had a race condition they couldn’t figure out. He hit on the idea of synchronizing the circuit to a clock. It was meant as a kludge until he had time to get back to debugging it. But he never did, and now nearly every computer in the world has a clock.

    Similar stories abound in the field. No one understood John McCarthy’s original notation for Lisp. He invented the syntax we now know as a way to explain the notation. To his chagrin, the AI world adopted it in lieu of his preferred syntax.

Submit comment

You must be logged in to post a comment.