1. Forum
  2. Usenet
  3. COMP.ARCH

  • Re: core memory, Historical evolution of CPU perf

    From John Levine@21:1/5 to All on Fri Oct 11 19:40:11 2024
    According to BGB <[email protected]>:
    We used the 11/20 as a remote debug device for the 8085 cash
    register machine(s) we were building.

    Core memory: slow to access but also slow to forget.

    Core memory was before my time, but I remember reading somewhere that it >needed to be preheated to a certain operating temperature in order to
    write to it (because the hysteresis energy of the ferrite rings was >temperature dependent, and it needed too much power to flip the bits at
    lower temperatures).

    The memory for the IBM 7090 in the late 1950s was in a heated oil bath
    but they soon figured out how to make core work at room temperature.
    On the PDP-8 and PDP-11's that I used, the core was in the box with
    everything else with no special temperature controls.

    --
    Regards,
    John Levine, [email protected], Primary Perpetrator of "The Internet for Dummies",
    Please consider the environment before reading this e-mail. https://jl.ly

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Stephen Fuld@21:1/5 to John Levine on Fri Oct 11 16:36:18 2024
    On 10/11/2024 12:40 PM, John Levine wrote:
    According to BGB <[email protected]>:
    We used the 11/20 as a remote debug device for the 8085 cash
    register machine(s) we were building.

    Core memory: slow to access but also slow to forget.

    Core memory was before my time, but I remember reading somewhere that it
    needed to be preheated to a certain operating temperature in order to
    write to it (because the hysteresis energy of the ferrite rings was
    temperature dependent, and it needed too much power to flip the bits at
    lower temperatures).

    The memory for the IBM 7090 in the late 1950s was in a heated oil bath
    but they soon figured out how to make core work at room temperature.
    On the PDP-8 and PDP-11's that I used, the core was in the box with everything else with no special temperature controls.


    And in the late 1960s and early 1970s the Univac 1108s and the IBM S/360
    models 30 and 65 that I worked with worked just fine in the typical
    cooled computer rooms of the day (IIRC high 60s degrees F)


    --
    - Stephen Fuld
    (e-mail address disguised to prevent spam)

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From John Levine@21:1/5 to All on Sat Oct 12 19:57:04 2024
    According to Scott Lurndal <[email protected]>:
    Compared to the alternatives at the time, it was the fastest
    gun in the west.

    At the time of 11/20 debute (1970) it already was not.

    "at the time" was in the late 1950s when it was first proposed,
    and through much if not all of the 1960s.

    Core memory was invented in the late 1940s. An Wang did well known work in 1949,
    and MIT's Whirlwind was the first working computer with core designed by Jay Forrester in 1953. I think the IBM 704 in 1954 was the first commercial system with core. By 1960 it had replaced all of its competitors including Williams tubes and delay lines.

    Core soon wasn't the fastest memory, which is why the Atlas and IBM 360/85 had caches in the mid 1960s. But it remained the best overall for a combination of speed, cost, and reliability for another decade.

    --
    Regards,
    John Levine, [email protected], Primary Perpetrator of "The Internet for Dummies",
    Please consider the environment before reading this e-mail. https://jl.ly

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