• CRISP/Hobbit (was: What is RISC?)

    From Anton Ertl@21:1/5 to Paul A. Clayton on Wed Dec 13 07:09:41 2023
    "Paul A. Clayton" <paaronclayton@gmail.com> writes:
    It looks like one can get a PDF from Semantic Scholar: >https://www.semanticscholar.org/paper/A-32-bit-processor-design-Johnson/5ef2b3e8a755a2c29833eba8ab61117c296d95ac

    The link to the PDF does not work for me. Anyway, looking at where it
    is coming from, this is likely to be a predecessor of the CRISP
    architecture. And indeed, in the CRISP paper <https://dl.acm.org/doi/pdf/10.1145/30350.30385> the authors say:

    |Since 1975 the Bell Labs C Machine Project has designed several
    |computer architectures to support efficiently the C Pro- gramming
    |Language. [2,4,5,6] These designs evolved into the current C Machine |instruction-set architecture.

    And [5] is Johnson's paper.

    CRISP has a memory-to-memory architecture. The basic idea of the C
    machine project seems to be to "close the semantic gap" with hardware,
    the typical justification of CISC features.

    CRISP eventually was commercialized as AT&T Hobbit <https://en.wikipedia.org/wiki/C-language_Reduced_Instruction_Set_Processor> (recommended), but there were no customers except AT&T-owned Eo (the
    hardware arm of Go). Reading up on <https://en.wikipedia.org/wiki/GO_Corporation>, AT&T seems to have
    been somewhat fickle-minded, researching and developing the C Machine
    project since 1975, then acquiring Eo in 1993 to finally have a
    customer for that, acquiring Go (the software for Eo) in 1994, but
    then canceling Hobbit two weeks later. I think this was the right
    decision, but they would have saved quite a bit of money if they had
    taken it earlier.

    - anton
    --
    'Anyone trying for "industrial quality" ISA should avoid undefined behavior.'
    Mitch Alsup, <c17fcd89-f024-40e7-a594-88a85ac10d20o@googlegroups.com>

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)