Digital Equipment Corporation PDP-7

Ken Olsen - 6th February 2011. We have recently received the sad news that Ken Olsen co-founder of Digital Equipment Corporation, has died. He was 84. Digital Equipment Corporation during the 1960's through to the 1980's was one of the major players in the computer industry with their range of PDP and VAX series computers. Ken died on Sunday, February 6, 2011. He would have been 85 years old on February 20. Our condolences to Ken's family.   More here, (copy).

PDP-7 sales photo

A PDP-7 in manufacture

PDP-7A S#115 undergoing restoration in Oslo Norway ©2009 Tore Sinding Bekkedal

1964-65 saw the delivery of the third design in Digital Equipment Corporation's series of 18-bit computers(1) - the PDP-7, the first of their computers to use "FlipChip" technology. A later PDP-7 revision, the PDP-7A, was produced using R series modules, and at the end of production a total of 99(2) PDP-7 and PDP-7A systems had apparently been shipped. As a minicomputer the PDP-7 had a cycle time of 1.75 microseconds(3) and add time of 4 microseconds, I/O included a keyboard, printer, paper-tape and dual transport DECtape drives (type 555). Of the original sales target of 120 PDP-7 systems, quite a number were used in laboratory and data acquisition applications. DEC provided an "advanced" Fortran II compiler, a Symbolic Assembler, Editor, DDT Debugging System, Maintenance routines and a library of arithmetic, utility and programming aids developed on the program-compatible PDP-4.

The PDP-7 was developed as a less expensive alternative to the earlier PDP range and it had an introductory price tag of only $72,000 for a minimal system configuration (options list). Utilising core memory as did most machines of it's era, the PDP-7's 18-bit memory started at 4K words upgradeable in 4k chunks to 64k. A range of I/O units was available including Calcomp plotters, DECtape magnetic tape, paper tape readers and high speed papertape punches, DEC 340 CRT display unit and the ubiquitous Teletype ASR33.

Of the 120 PDP-7's originally built only four are currently confirmed to exist, with two only being in some sort of operable condition.

1. #47, recently confirmed, is a minimal configuration system at a museum in Australia owned by Max Burnet, who has kindly provided photos and information about the system and DEC Australia in general.
2. #113, at the University of Oregon was operated by Dr. Harlan Lefevre until his retirement when it was decommissioned and recently donated to a museum, it is believed this machine was still fully operational after nearly 40 years with some 65,000 hours of use logged. It is apparently due to be recommissioned in 2009.
3. #115, is apparently undergoing restoration in Oslo, Norway by Tore Sinding Bekkedal, but it is believed this machine had most of its guts modified to TTL 74 series logic at some point, which was then recovered when the machine was decommissioned.
4. #33. Rumors of a fourth PDP-7 system at the Computer History Museum (Mountain View, California), have now been confirmed following an email from Len Shustek, chairman of the museum. The museums online entry for this machine is here. This machine is listed as being from Massachusetts General Hospital so would have been either PDP-7 S#33 or PDP-7A S#102, from the photos we suspect it is #33.

C.H.M. also have several photos of PDP-7 systems in manufacture, including two views of serial #47, the system mentioned above, that was shipped to the Australian Atomic Energy Commission.

The PDP-7's claim to fame was forged around 1969/70 when Ken Thompson used a scavenged PDP-7 at Bell Labs, the research arm of AT&T, to develop the operating system that later became Unix. (From the 18-bit Service list this would appear to have been one of PDP-7's #3, #34, #44 or PDP-7A S#149.

PDP is an abbreviation for Programmed Data Processor, coined by Digital Equipment Corporation and a Boston venture capital company American Research and Development. There was a perceived view at that time that with a limited world market for "Computers" of just a few hundred, of what were seemingly big and expensive machines requiring a dedicated computer center and a large supporting staff, the term "Computer" should be avoided as there was no money to be made in computers ! (doh!), and so was born the PDP - the "Programmed Data Processor".

Links -

Tore Bekkedal	Restoration of a PDP-7 in Oslo.
PDP-7 data	Possibly the largest archive of PDP-7 data available. and a mirror site here
Unix history	Origins and History of Unix, 1969-1995.
History	Unix - "The famous PDP-7 comes to the rescue".
Wikipedia	PDP-7 entry at Wikipedia.
PDP-7 Service list	Details of the known 99 PDP-7 systems built.
Ken Olson	Biography and the story of DEC.

If you know of any more information about the PDP-7, options, location of existing systems, spare parts, ancillary bits, software, tapes or manuals, then please let us know here.

Further reading on PDP-7 history and design can be found here.

Our own library of PDP-7 information can be found here.

PDP-7 Service list (1972)
The following list information was compiled from Digital Equipment Corporations 1972 18-bit Customer Service List (kindly supplied by Bob Supnik), available here (6.5Mb pdf), and lists the 99 known PDP-7 and PDP-7A systems on the list in 1972. 120 systems were forecast to be built in total, but at this time we do not have any further information about the possible remaining 21 systems or if they were even built. The PDP-7 appeared to have sold well into Government research and University sectors with 11 systems shipped to the UK alone, almost 10% of the forecast production run !. Serial numbers are concurrent for both PDP-7's and the PDP-7A's giving 102 systems shipped but with three missing in the list. The missing 21 could be of either type, however we are now confident that the 99 systems shipped were the only ones ever built.

Systems in bold are systems existing today in museums or private collections, although not necessarily in an operational state. One machine however S#113, has recently been restored and is now fully operational and open to public viewing at the Living Computer Museum in Seattle. The 18-bit Customer Service List covers PDP-7 S#1-50 with S#22 and S#46 missing and PDP-7A's S#101-152 with S#125 missing, making it 99 systems shipped of the potential 102 serial numbers.

A full list, to our knowledge, of the available options on a PDP-7 is available here.

Miscellaneous DEC information, manuals, data sheets, Etc. mostly for the PDP-7 of course !, but there is some PDP-11 and PDP-15 info in here as well.

1. DEC 18-bit Computers -
   PDP-1 - 4k-64k of main memory, 5uS cycle time, one's complement arithmetic. Mostly constructed of DEC 1000-series system building blocks rated at 5MHz. Apparently three PDP-1 computers still exist, all in the collection of the Computer History Museum, a prototype, and two PDP-1C machines. The last PDP-1 manufactured (#55) has been restored to working order and is used as a working exhibit.
   PDP-4 - First shipped in July 1962 as a slower, cheaper alternative to the PDP-1, but not commercially successful. One customer of these early PDP machines was Atomic Energy of Canada. The PDP-4 has been recently implemented by David Conroy in a XILINX field programmable gate array.
   PDP-7 - Replacement for the PDP-4, DEC's first wire-wrapped machine.
   PDP-9 - Successor to the PDP-7, DEC's first micro-programmed machine. This website by Bob Pooler is dedicated to the Engineering Techs, Engineers & Production staff of the PDP-7 & PDP-9 Systems Group.
   PDP-15 - DEC's final 18-bit machine. It was their only 18-bit machine constructed from TTL integrated circuits rather than discrete transistors. Later versions of the system were referred to as the "XVM" family. The new machine was faster and less expensive than its predecessors and had the added sophistication of a separate I/O processor to the CPU. Over 400 of these machines were ordered in the first eight months of production. PDP-15's in the collection of Mike ?
   (back)
2. Various information on the internet gives sales of the PDP-7 systems as 120 units, however we have not found any information to substantiate this number. The DEC historical document "Nineteen Fifty-Seven To The Present " lists on page six that 120 systems were sold, whilst on page 86 it shows 115 systems !. From the DEC book "Computer Engineering - A DEC view of hardware systems design", it is apparent that the original sales goal for the PDP-7 range was 120 units, but to date the only firm evidence for the number of systems produced is the 18-bit Service list, showing 99 systems. Unless further information surfaces in the future, which is probably unlikely, 99 shipped systems it will have to be.
   (back)
3. A recent find in DECuscope Volume 10 1971, the newsletter of the Digital Equipment Users Society, has unearthed an article by A. R. Atherton of the Cavendish Laboratory, Cambridge, England (PDP-7A S#126), entitled "MODERNIZING A PDP-7", where a new 16K core store was fitted to their PDP-7 and the opportunity taken to reduce the cycle time of the computer from 1750 nsec to 875 nsec, doubling its speed.
   (back)

Site design ©2007- www.soemtron.org / .com -   []