Apollo computer

Discussion in 'HardForum Tech News' started by $trapped, Jun 15, 2019.

  1. $trapped

    $trapped [H]Lite

    Messages:
    94
    Joined:
    Jan 18, 2012
    This story appeared recently in Fast Company as part of a series they are running on the moon landing. Sometimes it's good to revisit technology milestones to appreciate just what we have today.
     
    DejaWiz, AceGoober, Jon855 and 3 others like this.
  2. DrDoU

    DrDoU 2[H]4U

    Messages:
    2,293
    Joined:
    Jun 4, 2007
    That’s some archaic stuff. Even a caveman can use it.
     
  3. Burticus

    Burticus 2[H]4U

    Messages:
    3,871
    Joined:
    Nov 7, 2005
    I had a high school teacher in like 1990 that went on and on about how his Casio digital calculator watch had more horsepower and memory than the stuff they used for the Apollo missions. Actually kinda scary if you stop to think about it... so don't.

    edit - also don't think too hard about the computers that control our nuclear missile arsenal... all ancient and some using floppy disks. But hey at least they can't be hacked, we hope.
     
    Last edited: Jun 15, 2019
  4. erek

    erek 2[H]4U

    Messages:
    2,488
    Joined:
    Dec 19, 2005
    https://news.ycombinator.com/item?id=12043206

    The Kalman Filter was used the in the Apollo 11 Guidance Computer [0] (discussed in the past on HN [1]).As someone linked previously, here is a historical perspective [2], and a link to the actual state vector update computations [3].The AGC maintained the state vectors for t...
     
    AceGoober and Red Falcon like this.
  5. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    I know you are being facetious, but I doubt anyone under the age of 70 would even begin to know how to operate any of that equipment.
    Mathematicians, assembly language programmers, and engineers who studied and practiced with such systems in the 1950s and 1960s (maybe the 1970s) would know how to work, operate, and maintain such equipment.

    Just because it has a low processing and memory count doesn't make it "simple" or easy to use by any stretch.
    Credit and respect where it is due to those who worked on, and engineered, such systems.


    From the article:
    I work on "retro" and legacy systems from the 1980s-present, and the older the systems are, the more custom, proprietary, and difficult they are to use; I can only fathom how systems from the 1960s operated, especially without higher-level programming languages, and oh lord - microcode and assembly language, for the entire system! :eek:
    A modern smartphone has the processing power of an early 90s supercomputer, yet a toddler can operate it with ease - that doesn't make the toddler smart, it just means the technology is that simple to use now (props to the programmers and engineers who made it that simple to use).

    These 1960s computers had less processing power than an Arduino controller has now, but the engineering and resourcefulness to make such an insignificant amount of processing power assist humankind with going to the moon was truly incredible.
    This was much more of an undertaking than pressing "Install" to install one's Netflix app on their smartphone...


    https://en.wikipedia.org/wiki/Margaret_Hamilton_(software_engineer)

    Margaret Hamilton is a genius and a master programmer and systems engineer - I could only dream of one day being that incredible, if ever!

    margaret-hamilton-mit-apollo-code_0.jpg
    (those stacks are Margaret Hamilton's Apollo code for the lunar module and command module)

    Margaret Hamilton in 2009:
    Margaret Hamilton in 2017:


    http://news.mit.edu/2016/scene-at-mit-margaret-hamilton-apollo-code-0817

    https://qz.com/726338/the-code-that...-to-github-and-its-like-a-1960s-time-capsule/

    Again, credit where credit is due!
    Her Apollo 11 code is also publicly available on GitHub:

    https://github.com/chrislgarry/Apollo-11/

     
    Last edited: Jun 16, 2019
    illram, noko, Crackinjahcs and 19 others like this.
  6. r00k

    r00k 2[H]4U

    Messages:
    2,655
    Joined:
    Aug 24, 2004
    I had a guy come in to my shop several months ago who worked on several of the early NASA missions. He had programmed a number of those early systems in machine code etc. Im a bit fuzzy on it all now. He also talked about how he had dumpster dove behind Apple's first store and had recovered some of their discarded logic boards and figured out how to put them to use unofficially.

    I had no reason to doubt him at any rate, as he was old and spoke about it all in levels of detail that suggested he really knew his shit..
     
    jeffj7, RanceJustice, GoodBoy and 3 others like this.
  7. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    Yep, they are all using 8" floppy diskette drives - those were considered old and legacy in the early 1980s, and were phased out completely in the world by the late 1980s.
    They also utilize 24v lines instead of 12v lines due to the large motors required to spin up the diskette.

    I know how to use 8" FDDs on semi-modern systems, and have a self-assembled kit and 24v PSU with molex connectors for them with one of D Bit's 50-pin Shugart SA800 to 34-pin FDD adapters - I won't be trying to hack any nuclear silo systems any time soon, though. ;)
    Props to D Bit for making the converter PCBs for them!

    http://www.dbit.com/fdadap.html

    fdadap.jpg

    Also helpful for making a cable adapter: http://www.classiccmp.org/dunfield/img54306/cnct.htm
     
    Last edited: Jun 16, 2019
  8. THRESHIN

    THRESHIN 2[H]4U

    Messages:
    2,951
    Joined:
    Sep 29, 2002
    I always found it fascinating. With the space program (same as nuclear technology at the time) it really is amazing that humans were able to pull it off. Our technology sucked and we were really pushing the boundaries of what it was capable of.

    Also interesting, during the landing of the LEM during Apollo 11, the guidance computer failed. Apparently there was too much data for the computer to handle! Armstrong and Aldrin made the decision to land it manually knowing that if they used too much fuel they would be unable to return to the command module and would die on the moon. Thankfully, they were able to do it and everything went fine after that. But I did read that it was barely....
     
  9. defaultluser

    defaultluser [H]ardForum Junkie

    Messages:
    12,364
    Joined:
    Jan 14, 2006
    Yeah, everything used in the Apollo Computer was cutting-edge design. Not just the transistors, but also the algorithms.

    In addition to the new math to put us into orbit, you also need an algorithm to filter potentially bad sensor updates.

    After Apollo, the filter ended up used in a lot of applications. Who knew that finding the good data (truth estimate) in a sea of ambiguities had such wide use? :D

    https://www.eng.ufl.edu/newengineer/in-memoriam/remembering-rudolf-e-kalman-1930-2016/
     
    Last edited: Jun 16, 2019
  10. Axdrenalin

    Axdrenalin [H]ard|DCer of the Month - Nov. 2009

    Messages:
    6,174
    Joined:
    Jan 28, 2004
    When I worked at Ft. Huachuca in the early 90’s, we actually still had some Sperry Univac hardware in the software development lab that used 8” floppies. Of course, we also had the archaic removable hard drive cabinets and platters as well and 9 track tapes to content with too, so I guess it shouldn’t be surprising that 8” FDDs were still a thing there. Looked like the old NASA data control centers from the 70’s. Left in ‘94 and never looked back.
     
    termite, GoldenTiger and Red Falcon like this.
  11. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    I'm really impressed much of that equipment was still in use into the mid-1990s.
    Did you have a chance to work first-hand with any of that equipment at all?

    I believe you can even get those top-loader HDD platters from here: http://www.athana.com/rigid/rd.html

    CT%20Pack1.jpg


    Also, just for kicks - email in 1984:

    (How many people under the age of 35 even know what a rotary phone is, let alone how to use one - I do! :D )
     
  12. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    Since we're on the topic, you all might appreciate a few of these systems as well from our past! (y)









    I can't recommend LGR's channel enough.
     
  13. lostin3d

    lostin3d [H]ard|Gawd

    Messages:
    1,995
    Joined:
    Oct 13, 2016
    Thanks for the post. Love these kind of stories.
     
    Red Falcon likes this.
  14. Axdrenalin

    Axdrenalin [H]ard|DCer of the Month - Nov. 2009

    Messages:
    6,174
    Joined:
    Jan 28, 2004
    I was there primarily for the Bytex autoswitch installation, implementation and support, but received a bit of cross training on the 9000 series AMME hardware they had in the lab. Looks like most that of that stuff was installed starting in 1974 in several military comm centers, and the SDL at Huachuca was probably one of the last remaining installations. I’m sure that what I was doing there with the Bytex was probably related to the eventual replacement of some of that old hardware.

    I found this article about the Univac Hardware on this site...looks even older than some of the stuff we used. Forgot about the card reader even... :)
    http://www.onegentleman.biz/Univac History/e and the Univac Computer.html

    Dang, I didn’t think it would be so hard to find reference material for the Bytex hardware! Wish I had some pictures from back in the day, but you know how it goes with military comm sites and security. This thread is definitely taking me back in time as well...
     
    Last edited: Jun 17, 2019
    AceGoober, Red Falcon and GoldenTiger like this.
  15. ZodaEX

    ZodaEX 2[H]4U

    Messages:
    3,722
    Joined:
    Sep 17, 2004
    Double post. We have a modern feature called "edit" for such situations.
     
    AceGoober likes this.
  16. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    Asshole post. We have a modern feature called "ignore" for such situations.
     
    Last edited: Jun 17, 2019
  17. ZodaEX

    ZodaEX 2[H]4U

    Messages:
    3,722
    Joined:
    Sep 17, 2004
    Actually take the time to read my post and your question will be answered.
     
  18. Verge

    Verge [H]ardness Supreme

    Messages:
    6,176
    Joined:
    May 27, 2001
    wish it had more technical details in it
     
  19. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    I didn't have a question, I know what you are saying, and your post was a complete waste of everyone's time here since you contributed absolutely nothing to this thread.
    You aren't a mod, so next time, please keep your shit posts to yourself.
     
    wolfofone likes this.
  20. ZodaEX

    ZodaEX 2[H]4U

    Messages:
    3,722
    Joined:
    Sep 17, 2004
    If you know what I'm saying, and I know what your saying, then what's the problem? This is a discussion board and we are discussing things. Grow up a little and quit whining constantly.
     
  21. Ebernanut

    Ebernanut Gawd

    Messages:
    998
    Joined:
    Dec 15, 2010
    I always enjoy reading about old tech like this, it helps illustrate how far tech has come when you read about all of the limitations they had to overcome back then.

    It's interesting to read that they allowed the programmers so much trust and freedom to do what needed to be done given how careful they are these days for fear of introducing hardware or software bugs. I guess since there was no older stuff to fall back on they just had to hope no errors made it through.

    Lol, do you not see the irony in this statement?

    I enjoyed the posts they had made, yours not so much...
     
  22. defaultluser

    defaultluser [H]ardForum Junkie

    Messages:
    12,364
    Joined:
    Jan 14, 2006
    They had system design by more than just a single person, test setups to evaluate the completed systems, and code reviews, just like we do today. See here:

    http://www.drdobbs.com/architecture-and-design/one-giant-leap-the-apollo-guidance-compu/184404139

    The reason you think it was just one engineer is because that's all they write the stories about. It was an effort that took hundreds of people to accomplish, even if the brains running the thing were fewer in number.

    Computers had already been sold by IBM and other big makers for fifteen years, so process was already well-defined, even if you saw a lot of rockstars that stood-out at the top.
     
    Last edited: Jun 18, 2019
  23. Ebernanut

    Ebernanut Gawd

    Messages:
    998
    Joined:
    Dec 15, 2010
    I wasn't trying to say I thought it was one engineer and I'm sure they still had a strict review process but I was referring to this quote above: "Because software was a mystery, a black box, upper management gave us total freedom and trust." These days NASA has a lot of bureaucratic hurdles for any new hardware and software, trust is not in their vocabulary.
     
    wolfofone likes this.
  24. defaultluser

    defaultluser [H]ardForum Junkie

    Messages:
    12,364
    Joined:
    Jan 14, 2006
    The reason they had to have total trust is because the algorithms were all new. You can't be choosy when there's only one algorithm that does the job, coded by this one person. Time and storage space are at a premium, so you're limited on your approaches.

    Once you leap over as many hurdles as Nasa has (50 years of them) and build up a long list of algorithms you have enough experience to fear change, and think you know everything.

    But that's not to say that, just because they depended on rockstars that there was no reviews; you had to prove your algorithm worked. They're just saying that "compared to NASA of today, they were completely on our side."
     
    Last edited: Jun 18, 2019
    wolfofone likes this.
  25. Ebernanut

    Ebernanut Gawd

    Messages:
    998
    Joined:
    Dec 15, 2010
    It's logical it's just so different from the way they've operated for a long time. I'm used to hearing about all the technical limitations they have to deal with due to their (justifiable) reluctance to adopt new tech that might have bugs.
     
  26. cjcox

    cjcox [H]ard|Gawd

    Messages:
    1,133
    Joined:
    Jun 7, 2004
  27. Dr. Righteous

    Dr. Righteous 2[H]4U

    Messages:
    3,124
    Joined:
    Aug 1, 2007
    Yeah, I found description of programming and problem solving for the Apollo program very compelling. No doubt, genius level intellect. This is what probably drew her to programming in the first place.
    All good until she got political in what she said. "Women are not allowed to drive in some countries. And women are not allowed to be priests" Yeah I'm sure "The Man" kept her down and kept her from her dream working for NASA and programming the Apollo computer. :rolleyes:

    When I was young I use to think as a rule women really were not as smart as men when it comes to technical things. I tried to learn assembler for my Atari 400 computer; because Basic was way, way too slow for a lot of things. Assembler kicked my ass. Couldn't handle it. I later read an article about programming of arcade machine games; all done in assembler and much by women. Whhaa??
    Facts are for women that find programming appealing their brain is more suited for "multitasking" and seeing interconnected big picture things. Men's thinking tend to be compartmentalized. It is like the analogy we often hear "Men are like waffles, women are like spaghetti." Men compartmentalize things and they are separate Cause and Effect situations. Thus a solution is narrow and plain to see to each. Women see everything as interconnected so a solution isn't always a solution since there are positives and negatives in different areas. When you are married you see this exact thing play out much of the time. You are presented with a problem and you know the solution instantly. But the wife will bring up possible repercussions of such a solution you didn't even consider and probably wouldn't have.
     
    Red Falcon likes this.
  28. Crackinjahcs

    Crackinjahcs Limp Gawd

    Messages:
    146
    Joined:
    Jan 31, 2018
    The joy of working in a results-based system, which can be hard to find nowadays.
     
    GoldenTiger, Ebernanut and Red Falcon like this.
  29. motomonkey

    motomonkey [H]ard|Gawd

    Messages:
    1,419
    Joined:
    Jan 17, 2009
    I was in the Army in the 80's and 90's, I can still remember our parts clerks lugging those huge TRS-80's out to the field, in the late 1980's, the best in semi-portable computing! those things were robust, all we had for power were small 60Hz utility generators, Voltage and Frequency would wander all over the place, with no power conditioners at all.
     
  30. sfsuphysics

    sfsuphysics I don't get it

    Messages:
    13,684
    Joined:
    Jan 14, 2007
    Kind of says something about the level of computing power that is actually needed for space missions, which really is not much. Hell most of the stuff in space is clunky old mostly because it's clunky durable they want it to keep working for years and years.

    And to go with a comparison, something like an Nvidia 8800 has more processing power than THE top super computers in the mid90s, however your phone still doesn't get anywhere close (regardless of how fancy it seems). Don't worry thought the fastest super computers in the early part of 2000s still are more powerful than even a 2080 TI :D Super computers simply made crazy strides from year to year, like some scary alien AI shit going to work.
     
  31. GoodBoy

    GoodBoy [H]ard|Gawd

    Messages:
    1,332
    Joined:
    Nov 29, 2004
    That's what happened, it's just not generally well-known by the public.
     
  32. MisterClean

    MisterClean 2[H]4U

    Messages:
    2,129
    Joined:
    Jun 9, 2010
    clockdogg and GoodBoy like this.
  33. Red Falcon

    Red Falcon [H]ardForum Junkie

    Messages:
    9,871
    Joined:
    May 7, 2007
    This is pretty cool and talks about Core Rope Memory that was used in the Apollo AGC:

     
    motomonkey likes this.
Tags: