The FP6000 - The Root of the 1900 Family Tree

 

 

 

It all began in Canada, Toronto to be more precise.

Here are a number of documents from the very earliest days of the FP6000 which would evolve to become the foundation of the 1900 Range.

    Filename Donor Description
    FP6000_manual ZS FP6000 Early description/brochure
    Q.151 - FP6000_Technical_Manual RDM FP6000 Technical Manual (Partial)
    Q.152 - FP6000 Fortran VKT FP6000 FORTRAN Manual (Includes 7 FORTRAN "Memos")
    I.M 25 MK I.M.25 - An early I.C.T. Document describing the workings of the 1904/5/9.

    The second part of this document describes the microprogram and should be used in conjunction with a large flowchart, which is currently not available.

    This I.C.T. document refers to a number of peripherals that were either never released, or were only used by one or two sites. (e.g. GD1830 was built, and possibly the GD1831 was something called the IDC.)

    (Please use Adobe to view as Edge thinks that it is on A8 paper or something microscopic.)

     

    The following are set of "FORTRAN Notes" used by the team developing the FP6000 FORTRAN II compilers and later in I.C.T. to develop the FORTRAN IV compilers.

    Fortran Note Donor Description of Content
    FN 14 VKT A quite detailed description of the process of compiling an expression.


    The following tables hold a collection of files, variously called 'Program Reports' (PRnnn), 'Project Memoranda' (PMnn) etc.


    Filename Donor Description of Content
    PM10 RDM Description of FP6000 Peripheral Control.

    This was FPCs 'standard interface' - 24-bit parallel interface, which had some advantages and disadvantages when compared to the I.C.T. Standard interface which eventually completely replaced this one.
    The biggest disadvantage was cost. Being 24-bit parallel and that each channel's controller had to occupy space within the CPU. The ICT 6-bit parallel interface was more appropriate for 'slow' peripherals like paper tape readers, card readers and line printers. The ICT interface was enhanced with a 'word' mode (24 bit) bit still running over 6-bit, but with faster timing. Later still a 'high speed' mode was built on that running bursts of 8 words at a time. The I.C.T. 1904/5/6/7 machines supported the FP6000 interface to allow the fast 1974 7-track magtape decks to be connected.

    PM22 RDM System Specification for FP6000 Drum Control
    PM23 RDM System Specification for FP6000 Magnetic Tape Channel Control (FP6501) and Magnetic Tape Unit Control (FP6520)
    PM24 RDM System Specification for FP6000 Card Reader Control (FP6300) and Code Converter (FP6301).
    PM28 RDM Peripheral Designations FP6000 -4, -5, -6, -7, -8
    PM29 RDM Notes on the FP6000 Order code.

    This is the only document that clearly documents the NORM (114 and 115) instructions.

    PM30 RDM Specification of Floating Point Arithmetic Unit (Part of FP6002 and FP6004).
    This should be read in conjunction with IM47.
    IM47 BitSavers I.C.T. Detailed document on the 1905 FPU, likely the FP6000 FPU, see PM30 above..
     I.C.T. Standard Interface (djvu)
     I.C.T. Standard Interface (pdf)
    KC I.C.T. Standard Interface Specification.

     

    Here are the available FPC 'Program Report' files.

    Program Report Donor Description of Content

    PR105

    RDM
    General Purpose Loader.

    On both the FP6000 and much of the following 1900 machines, programs were loaded by executive using a multiple stage process.
    1. Exec would read the first CARD or Paper Tape record and determine the amount of core required to hold the loaded program. Aborting if not enough free to hold program.
    2. Exec would copy a small program, whose image was a part of exec into the bottom of the area allocated. It would then 'seed' this image with the name of the program to be loaded and start it running.

    3. This small user program, the General Purpose Loader (GPL) would then use the device it was told to use by exec to load the image into core. This image would have been in what was called 'Consolidated Semi-Compiled' which was basically a concatenation of the semi-compiled versions of the program, followed by the semi-compiled portions of any required runtime libraries. - i.e. The GPL was doing what the later consolidators programs did when writing the final program image to tape or disc.
    There appears to have been a number of different GPLs, one for each medium type used to store programs (PT, CD, MT or DA). A re-created source of the punched card GPL has been created and is available here.
    PR106 RDM
    A description of program #CNSL, the FP6000 Consolidator.

    This worked by processing the paper tape semicompiled output by the early compilers and in conjunction with any required runtime libraries etc, it would emit the 'Consolidated Semi Compiled' that would then be loadable using the GPLs described above. Anyone who wants to see inside an actual consolidator, see CRCC. - a re-creation of the #XPCC card consolidator.
    Last updated: 23-Feb-2026