|
Introduction
List of Known Codes
Further Reading
Content by Tomáš Slavotínek, Tim Clarke, and Christian PL. Last update: 27 Dec 2023
Introduction
IBM plant codes are 2-digit numbers identifying the facility where a
given product was manufactured and/or assembled. These codes can be found on
many IBM products - most significantly on assembled units and integrated
circuits.
Systems and Other Units
Main units, monitors, terminals, storage units, and various other IBM
products typically come with a type/serial number sticker. The unique serial
number often starts with a 2-digit prefix - the plant code. Something like
this:
S/N 55-2Z0R2
Here the "55" indicates that the machine (IBM PS/2 Model 95) was
manufactured/assembled in Greenock, UK.
Integrated Circuits
Similar information can be found on many IBM-made ICs.
However, our understanding of the used format and values is very limited at the
moment. In general, the code consists of an "IBM" prefix, followed by one or
two 2-digit plant codes, terminated by an optional "Q", "P", or "PQ" suffix.
Here are a few examples:
IBM 53 52Q
IBM 93 14
IBM 9352PQ
IBM 9352
IBM 98
IBM98
The spacing between the individual sections seems to be greatly inconsistent
and most likely can be ignored.
The "IBM" prefix is self-explanatory. Sometimes it appears in form of the
striped IBM logo, but often uses a more pedestrian font shared with the rest of
the chip labeling.
It's currently unknown what the "Q"/"PQ" suffix indicates. The most common
theory is that it stands for "Qualification" and "Pre-Qualification"
respectively. But if that's the case, then IBM shipped an awful amount of
pre-qualification IC samples in their products - which seems rather
questionable. The rare "P" suffix might stand for "Prototype" (seen on
THIS unreleased adapter).
Some chips have a single 2-digit plant number, while others have two 2-digit
numbers (4 digits total). The latter variant possibly indicates where the die
was fabricated and where the chip was packaged - in cases where the two aren't
the same. If this is correct it seems more likely that the first two digits
would indicate the assembly line and the other two the silicon fab.
List of Known Codes
Some of the information presented here is currently unverified and may
therefore be inaccurate.
Duplicate/conflicting information is highlighted.
| # | Location | Product | Notes | Source |
|---|
| 00 | ("USA generic"?) | | | |
| 01 | Endicott, NY, USA | S/370 and many others | IBM's first manufacturing facility; see HERE | 4 |
| 02 | Poughkeepsie, NY, USA | | | 1,4 |
| 02 | Belgium | | | 1 |
| 06 | China | | | 4 |
| 08 | Beaverton, OR, USA | | | 4 |
| 08 | Netherlands | | | 1 |
| 10 | Rochester, MN, USA | S/3, S/32, S/34, S/36, S/38, AS/400 | see HERE | 1,4 |
| 11 | Lexington, KY, USA | | | 1 |
| 13 | San Jose, CA, USA | | | 4 |
| 14 | (Burlington, VT, USA?) | IC | very common; CMOS?; often as "5314", "9314" | |
| 16 | Sweden | | Vällingby (punchcards) or Järfälla (printers)? | 1 |
| 16 | ? | IC | possibly recycled code? seen on T3 complex | |
| 17 | Switzerland | | | 1 |
| 21 | Guadalajara, Mexico | Power systems | | 4 |
| 22 | ? | IC | | |
| 23 | Boca Raton, FL, USA | PC, PS/2, and others | Entry Systems Division (ESD); until 1996 | 3 |
| 23 | Raleigh, NC, USA | PC, PS/2, and others | | 4 |
| 26 | Austin, TX, USA | | | 1 |
| 27 | ("3rd party"?) | units | seen on the "Made for IBM" 7690 | |
| 31 | United Kingdom | | | 1 |
| 34 | ? | IC | | |
| 35 | France (Bordeaux?) | | | 1 |
| 40 | Germany | | | 1 |
| 41 | Charlotte, NC, USA | | | 4 |
| 43 | Italy (Vimercate?) | | | 1 |
| 44 | Santa Palomba, Italy | S/36 and others | seen on many 5363 | 1,2,4 |
| 51 | Montpellier, France | IC | | 1,4 |
| 52 | (East Fishkill, NY, USA?) | IC | mfg/pkg; very common and long-lived | |
| 53 | (USA?) | IC | mfg/pkg; possibly also Fishkill? | |
| 54 | Havant, UK | | | |
| 55 | Greenock, Scotland, UK | PS/2 and others | | 2 |
| 58 | Amsterdam, Netherlands | | | 1 |
| 65 | Dublin, Ireland 93b | i, p systems? | | 1,4 |
| 68 | Singapore | | | 4 |
| 70 | ? | IC | always as "7070"? | |
| 71 | Germany | | | 1 |
| 72 | Raleigh, NC, USA | PC, PS/2, and others | Research Triangle Park (RTP); 1996 and later | 3 |
| 74 | ? | IC | | |
| 75 | Mainz, Germany | | | 4 |
| 75 | Vác, Hungary | Storage Products | | 4 |
| 77 | Spain (Valencia?) | units | | |
| 78 | Guadalajara, Mexico | | | 1,4 |
| 82 | Sumaré, Brazil | units | | 4 |
| 83 | Dublin, Ireland 99b | z, p systems? | | 1,4 |
| 84 | Singapore | | | 4 |
| 90 | Wangaratta, Australia | | | 1 |
| 91 | Don Mills, Canada | | seen in a 5291 terminal | 1,3 |
| 92 | Canada | | | 1 |
| 93 | Canada (Bromont?) | IC | mfg/pkg/test?; see HERE and HERE | 1 |
| 97 | Fujisawa, Japan | PS/55 and others | | 2,4 |
| 98 | Japan | IC | mfg/pkg; mostly PQFP, some metal can PGA | 2 |
| 694 | United Kingdom | | 3-digit code? | 1 |
Sources
- IBM typewriter reference (thx to Tim Clarke) — see HERE
- Empirical evidence (system units of known origins, etc.)
- Confirmed by Joe George (worked for an IBM dealer of midrange systems)
- Posted by Michael Quaranta (IBM Technology Expert Labs) — see HERE
Notes
- Plant codes "53 14" and "93 14" appear on many PowerPC 601(+) CPUs. These
chips were produced only at Burlington, VT, USA and East Fishkill, NY, USA.
It's however unknown how this maps to the plant codes. The same code
combinations can be found on other (CMOS?) chips from the same era.
- From an ex-IBMer:
I worked for IBM for 25 years (from '88-'13). I worked in Burlington, VT in
the chip design area. It's likely the IBM chips you see in this would have been
made in East Fishkill, NY, which handled the bipolar chips used by the IBM
Server group, whereas Burlington handled CMOS chips, more commonly used in
IBM's consumer and low-end products like the PC division, at least until the
mid 90's when Fishkill eventually switched over to CMOS. [...]
- In 1996 IBM closed Boca Raton manufacturing (Entry Systems Division) and
moved it to Raleigh (Research Triangle Park).
Further Reading
|
