Mac Pro 1965

Discussion in 'Mac Pro' started by hugodrax, Oct 26, 2013.

  1. hugodrax, Oct 26, 2013
    Last edited: Oct 26, 2013

    hugodrax macrumors 6502a

    Joined:
    Jul 15, 2007
    #1
    This is the CDC 8600 that was being worked on during 1968-72 until Seymour Cray left to form Cray and the work on the Cray-1 super computer.

    Amazing the similarities.

    http://en.wikipedia.org/wiki/CDC_8600

    [​IMG]
     
  2. DanielCoffey macrumors 65816

    DanielCoffey

    Joined:
    Nov 15, 2010
    Location:
    Edinburgh, UK
  3. gpzjock, Oct 26, 2013
    Last edited: Oct 26, 2013

    gpzjock macrumors 6502a

    Joined:
    May 4, 2009
    #3
    Crays were supercomputers that defined supercomputing. Check this behemoth with integrated seating from the 70's: http://commons.wikimedia.org/wiki/File:Cray_X-MP.jpg

    $5-8 million each.....with Freon refrigeration to help it reach the dizzying speeds of 80 MFLOPS without melting.
     
  4. fastlanephil macrumors 6502a

    fastlanephil

    Joined:
    Nov 17, 2007
    #4
    I wonder how it compares to the nMac Pro as far as speed.
     
  5. hugodrax thread starter macrumors 6502a

    Joined:
    Jul 15, 2007
    #5
    Someone should create a "Cray" platform you can sit your new Mac pro on, so it looks like a mini cray. and maybe put some ports on it or maybe inside the platform an SSD slot :)


    Anyhow the early Crays were so sensitive to the speed it took to send the data from one end of the cable to the next, that he had to create delay circuits to slow down the speed on the shorter cable runs to match the longer ones so that all the signaling/data would arrive synchronized together.
     
  6. Beardy man macrumors regular

    Joined:
    Dec 4, 2007
    #6
    ...or get one of those 'Steve' figures and sit it on a seat around the nMP. Sounds like it should work? :)
     
  7. ScottishCaptain macrumors 6502a

    Joined:
    Oct 4, 2008
    #7
    Okay, let me clear a few things up here since mainframe history is a bit of a hobby for me (and I happen to have an S/390 MP9K sitting in my basement).

    1) Mainframes from back then share absolutely nothing in common with modern day technology, except for the fact that some of the theories (not implementations) evolved into research and development that spawned the stuff we've got today.

    2) Modern day computers don't require an auxiliary HMC or service workstation to boot up ("boot up" ~= "IPL" in the mainframe world). Virtually no mainframe ever produced was capable of self booting without the assistance of an external computer that initialized the processor complex, memory subsystems, loaded the initial program and enabled execution. The SMC chip in a modern day Mac Pro bears absolutely no resemblance at all to the service elements of a mainframe.

    3) If mainframes and personal computers were so similar, IBM wouldn't still be selling their zSeries mainframes today for exorbitant amounts of money and runtime licensing. If you need to run something that deals with high volumes of data and cannot fail, you run it on a Mainframe- not a Mac Pro.

    TLDR; there are no real similarities between a Mac Pro and a CDC. If you're really interested in reading up on this stuff, Andras Tantos wrote an excellent Cray X-MP simulator and a series of articles detailing his endeavours to emulate this system. That should give you an idea of just how insanely different the personal computer and mainframe markets were (and still are).

    http://modularcircuits.tantosonline.com/blog/articles/the-cray-files/

    -SC
     
  8. ScottishCaptain macrumors 6502a

    Joined:
    Oct 4, 2008
    #8
    Say what?

    Cray was obsessed over wire speed so much that this alone basically gave birth to his iconic torus design(s). High speed components were strategically placed within the machine to minimize cable length, and the wiring they used to hook everything up was all cut to a precise length to avoid the exact issue you're talking about.

    There were no delay circuits. That would have been suicide in terms of performance. So I'm not sure what you're talking about.

    -SC
     
  9. AidenShaw, Oct 26, 2013
    Last edited: Oct 26, 2013

    AidenShaw macrumors P6

    AidenShaw

    Joined:
    Feb 8, 2003
    Location:
    The Peninsula
    #9
    This is exactly why PCI Express replaced PCI, why SAS replaced SCSI, and why SATA replaced PATA.


    A common tactic to deal with wire speed delays was to make all the wires the same length - even if a shorter wire would reach from point A to point B for some of the bits.

    On PCBs sometimes you can see that the "shorter" traces are run on zig-zag paths to make them the same length as the longer traces.
     
  10. ScottishCaptain macrumors 6502a

    Joined:
    Oct 4, 2008
    #10
    Yeah, that's what I was saying.

    The early Cray machines were basically wired up by hand, which is why you always see a mass of wiring on one side of those machines (typically on the inside of the torus) and why there always seems to be an excess of wiring- it's because each wire within each cluster of wires were all the same length (and beyond that, each cluster of wires within each layer of the torus were all typically the same length as well), irregardless of the minimum physical length required to go from point A to point B.

    At the same time, back then longer wires meant slower speed- so the whole point of Cray's famous torus design was to shorten the distance of wiring between components, for which the fastest components were positioned near the centre of the torus (decreasing the collective wire length) and the slower components placed furthest from the centre of the torus (where a slightly longer wire length didn't matter as much).

    Which is why his statement puzzles me. To the best of my knowledge, there were no delay circuits- anything of the kind would absolutely murder your performance. That issue was solved by using precise wire lengths, and the machine was further accelerated by bending the entire configuration into a torus.

    Even in modern day electronics, as you point out- they keep the trace lengths the same to avoid this issue. I've never, ever heard of or seen a "delay circuit" being used to solve the issue of a signal propagation delay being unequal across a bus. I can't even begin to imagine how complex and quirky that kind of setup would be, especially when the solution is so simple (keep the trace lengths the same).

    -SC
     
  11. shiekh macrumors member

    Joined:
    Sep 26, 2006
    #11
    On modern RAM, some tracks switch back for the exact same reason.
     

Share This Page