The pictures at the "nfrpartners" address belong to John Coelho. I hope they stay available there.
There is a longer list of number in this file which I am still studying.
Finally, this is a list from IBM Brazil.
The 026 was used in the 70x and 70xx era.
The 029 replaced the 026 in the S/360 era. Photo of keyboard at http://www.nfrpartners.com/comphistory/029.htm and of card decks at .../punchcards1.htm
I (Lars Poulsen) used these in 1970 at Copenhagen University.
650 cards per minute, mechanical feed, regular card hopper, 13 pockets.
1000 cards per minute, mechanical feed, regular card hopper, 13 pockets.
2000 cards per minute, vacuum feed, long card feed ramp.
Used as an offline printer for the IBM 1620 computer.
Main storage was a drum, each instruction contained the address to jump to for the next instruction. The SOAP optimizing assembler took care of placing the program so that the next instruction would be ready to read just in time (accounting for the varying amounts of time each instruction took to execute).
A good reference is IBM 650 Technical Fact Sheet from IBM website
The 709 was a vacuum-tube realization of the same architecture that was later built in transistors as the 7090.
"The 1012 was impressively overbuilt. It was a reel-to-reel punch in a
cabinet about the size of a medium refrigerator. There were tension
arms on both sides between the supply and takeup reels and the
punch/read heads. Punch speed was 150 cps. The device was acquired for
an application that regularly required punching dozens of reels of
aluminized mylar tape. Prior tests showed that other punches didn't
endure this gracefully."
"The 1012 was impressively overbuilt. It was a reel-to-reel punch in a cabinet about the size of a medium refrigerator. There were tension arms on both sides between the supply and takeup reels and the punch/read heads. Punch speed was 150 cps. The device was acquired for an application that regularly required punching dozens of reels of aluminized mylar tape. Prior tests showed that other punches didn't endure this gracefully."
Photos at http://www.nfrpartners.com/comphistory/1052a.htm and .../1052b.htm
I used this in 1970 at Copenhagen University.
The entire Functional Characteristics Manual is at Howard Shubs' Website. Photos at http://www.nfrpartners.com/comphistory/1130a.htm and .../1130b.htm
I used this in 1970 at Copenhagen University.
Photo at http://www.nfrpartners.com/comphistory/2311a.htm
Long feed ramp. Looked a lot like the 2540, which was the corresponding unit for S/360. Like the 2540, it had 5 stacker pockets: The middle one could be addressed both by the reader and by the punch.
Photos at http://www.nfrpartners.com/comphistory/1403a.htm and .../1403b.htm
The 1403 was available in several models; the 1130 system could use a simpler 600 lpm version, while the System/360 family used the full-featured 1100 lpm model N1 shown in these pictures..
I saw this in 1970 at NEUCC, Copenhagen.
Photos at http://www.nfrpartners.com/comphistory/1419a.htm and http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig003.jpg and .../c20-1684/fig096.jpg
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig098.jpg
There was a CDC 1620 - was there also an IBM 1620?
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig013.jpg
I saw a 2250 in 1971 at NEUCC, Copenhagen.
The cluster controller for the 2260 dilpsays was the 2848.
I saw 2260s in 1971 at NEUCC, Copenhagen, where they were used as auxiliary consoles for the 360/65 in user areas.
Photos at http://www.nfrpartners.com/comphistory/2311a.htm and http://www.beagle-ears.com/lars/engineer/comphistory/c20-1684/fig054.jpg
I saw a bunch of these on the 360/65 at NEUCC, Copenhagen in 1971. Photo at http://www.beagle-ears.com/lars/engineer/comphistory/c20-1684/fig127c.jpg
The bins (sub-cells) "made a very good vase for long-stemmed flowers". "A device that Rube Goldberg would have declared to be far too complex". According to a former IBM CE (Customer Engineer) the strip accellerates to 1200 RPM in 1/8 of a revolution as it is pulled out of the cell and it can be read on the first revolution. The mechanism is extremely finicky, and any mishandling by a service engineer results in catastrophic failures: "Mechanical malfunctions "cascade', bits and pieces of assorted mechanism get strewed about the interior which is [otherwise] tidy enough to use a clean room filter". (firstname.lastname@example.org, alt.folklore.computer 2000-08-13) "As for the Data Cell, without its overtime, I would not be able to have near the toys that I have in retirement." (ditto, 2000-08-09)
Photos at http://www.nfrpartners.com/comphistory/2321al.htm and http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig055.jpg and .../c20-1684/fig056.jpg
The motivation for building this monster was that disk drives were so expensive that putting 30 million drivers' license records on a hard disk was out of the question (and tapes were too slow for real-time lookups).
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig010.jpg
Photo at http://www.nfrpartners.com/comphistory/2400a.htm
We had two of these on an 1130 at Copenhagen University in 1970.
600 cpm and 1000 cpm versions. Used on S/360 and 1130. Photo at Howard Shubs' Website.
1000 cpm reading, 200 cpm punching. Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig011.jpg Note the long feed ramp. Same 5 stacker pockets as the 1402.
Often pronounced as the Mother-F*cking Card Mangler. Two input hoppers, 5 output stackers and a program-controlled card path between them.
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig089.jpg
The 2701 was the first TCU for the S/360. It supported only a few lines, but had an RPQ that allowed T1/E1 lines (up to 2MBps). Until the 1990s, this was the only IBM controller that supported T1 line speeds.
I wrote a terminal driver for using these under Univac EXEC-8 operating system at Copenhagen Unviersity in 1972, but we had already had the terminals for a couple of years before that.
Used to attach 2401 drives.
Used to attach 2303 drum, 2311 disk and 2321 Data Cell.
Used an acoustic delay line to hold the screen buffer. The character generator used a magnetic core ROM to hold the character dot matrix lookup table.
Photo at http://www.obsoletecomputermuseum.org/ibm3275.html
Prints on fanfold or roll stock and will optionally cut sheets. 30 inches of linear paper per second, or 3.5 pages per second (about 200 pages per minute).
A robotic tape storage system, featuring tape cartridges about the size and shape of a 12-ounce soda drink can, containing a wide strip of tape wound on a spool.
The cartridges were stored in two facing walls of honeycomb-arranged slots. Mechanical pickers (one or two, depending on the 3850 model) went back and forth between the storage walls, moved vertically and pivoted to reach the desired slot, then pulled a cartridge and carried it to one of multiple tape drive stations. At the drive, the cartridge cover was removed and the tape was read or written using helical-scan heads like those on a 4mm or 8mm digital tape drive. Each cartridge held about 50MB, a 3330 drive image filled two cartridges.
There were dedicated 3330 disk drives onto which data was staged from cartridges. Once staged, the data was accessed from the 370 mainframe like ordinary disk data. The entry-level 3850 had a total storage capacity of 35GB, but a fully-expanded system could hold much more (up to 472 GB). After the 3330 DASD went out of fashion, 3830 MSS systems were fitted with 3350 drives, but the 3830 code was never updated to use the additional capacity of the newer drives.
Pictures at http://www.coulmbia.edu/acis/history/mss.html.
Photo at http://www.obsoletecomputermuseum.org/ibm5100.html
Photo at http://www.obsoletecomputermuseum.org/ibmpc.html and at http://www.xs4all.nl/~rimmer/ibmpc.htm
Photo at http://www.obsoletecomputermuseum.org/ibm_port.html
Started out as IBM copier 3, to which a computer interface was added (and a magnetic card reader input device).
Lots of information at http://www.brouhaha.com/~eric/retrocomputing/ibm/stretch/
The 7090 was a transistor realization of the same architecture that was previously built in vacuum tubes as the 709.
Field Service did not like these at all.
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig077.jpg
Photo at http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig016.jpg
The 7950 Harvest combined a modified 7030 Stretch processor with a 7951 processsing unit.
A PC/AT with channel attach card and LAN cards, providing a mainframe gateway to T/R (token ring) and ethernet LANs. (Late 1980s)
$Log: ibm_nos.htm,v $ Revision 1.21 2004/01/23 16:36:32 lars *** empty log message *** Revision 1.20 2003/12/29 04:25:39 lars *** empty log message *** Revision 1.19 2003/11/21 15:31:05 lars Added 9672. Revision 1.18 2001/10/26 13:28:01 lars Replaced CMC -> Beagle-Ears Revision 1.17 2001/06/24 17:19:54 lars *** empty log message *** Revision 1.16 2001/04/08 16:39:53 lars Update links. Revision 1.15 2001/01/14 04:49:46 lars *** empty log message *** Revision 1.14 2001/01/08 05:32:55 lars Added a list from IBM Brazil. Revision 1.13 2001/01/05 06:37:39 lars More corrections from a.f.c readers Revision 1.12 2001/01/04 05:37:55 lars Yet more model numbers from a.f.c readers. Revision 1.11 2001/01/04 05:12:23 lars More models from an anonymous source. Revision 1.10 2001/01/04 05:00:56 lars Al Kossow provided many 70x and 70xx peripherals. Revision 1.9 2001/01/03 17:11:10 lars Added 27xx/3725. Revision 1.8 2000/11/22 08:00:38 lars *** empty log message *** Revision 1.7 2000/11/07 16:22:15 lars *** empty log message *** Revision 1.6 2000/10/02 06:15:32 lars Revision 1.5 2000/09/09 05:46:49 lars Revision 1.4 2000/08/28 06:35:01 lars Revision 1.3 2000/08/23 04:43:49 lars Revision 1.2 2000/08/15 01:25:07 lars