Locklin on science

The largest computer ever built

Posted in big machines by Scott Locklin on March 28, 2013

While cold war jets are an old interest of mine, almost everything built to fight the cold war fascinates me. All ages are characterized by madness; only a few have that madness captured in physical objects. Consider the largest computer ever built: the “Semi-Automatic Ground Environment” or SAGE system.

The SAGE system was designed to solve a data fusion problem. Radar installations across North America kept watch against Soviet bombers. These needed to be networked together and coordinated with air defense missiles and interceptors. Seems simple, right? In the 1950s and 1960s, this was not simple. The country is big; hundreds of radar stations and sensors needed to be integrated. It wasn’t as easy as it was in England in WW-2, when enemy aircraft location was plotted by hand on maps as the radar data came in: North America is much larger, and the planes traveled much faster in the 50s and 60s. No group of people could really make the decisions in time to mount an effective defense. You needed some kind of computer to make the decisions.

These giant electric brains took up an acre or so of real estate, and were encased in huge windowless concrete pillboxes all over the country.


The  SAGE system had many firsts: it was the first nation wide networked computer system. While it used special leased telephone lines and some of the first modems (at a blistering 1300 baud), it was effectively the internet, long before the internet. It was the first to use CRT screens. The first to use a “touch screen interface” via the use of light pens on the CRT. It was the first to use magnetic core memory. It was the first real time, high availability computer system. It was the first computer system to use time sharing. Many people attribute the genesis of computer programming as a profession to the SAGE system. Modern air traffic control, and computer booking systems of course, descend from the SAGE system.


Each of the 27 computers that made up the system was a dual core 32 bit CPU made  of 60,000 vacuum tubes 175,000 diodes, and 12,000 of those newfangled transistors. Memory: 256k of magnetic core RAM (invented for this project) clocking in every 6 microseconds. These things weighed 300 tons, consumed 3 megawatts of electrical power and ran a blistering 75,000 operations per second. The dual cores weren’t used for multiprocessing. One was kept on hot standby in case the other failed. Since our grandfathers knew about fault tolerance, they had a system to replace the tubes before they burned out: downtime was typically only a couple hours a year.


Each one drove 50 to 150 GUI workstations, and interacted with more than a hundred radars, interceptors and missile batteries. Remember that the next time you whine that your computer ain’t fast enough. No. This thing is less powerful than even a shitty cell phone (the 386 was probably approximately equivalent), and it did significantly more than your PC does. Each one was capable of coordinating the air defense of the entire North American continent.


It is worth pointing out that these machines not only ran all that equipment, and dealt with all that data, they also guided interceptors to their target locations. The F-106 and F-102 could be directly controlled by the SAGE system after takeoff. We think of “drones” as the latest and greatest newfangled thing in warfare: they have actually existed for a very long time. In many ways the SAGE system was more impressive than, say, the Predator system.


Another interesting piece of the SAGE system was the  BOMARC missile system.  The BOMARC (made by Boeing and the Michigan Aeronautic Research Center -which no longer exists) was primarily ramjet powered, and carried either a small nuke, or a half ton of conventional explosives. It was entirely dependent on the SAGE system for guidance to target. It was also incredibly stupid and dangerous: the original rocket boosters used hypergolic fuels, and would occasionally spectacularly explode in their silos, spreading dangerous plutonium around.

The SAGE system started running in 1958, and didn’t stop until 1984. Was it necessary? Like many interesting cold war artifacts, SAGE was more or less made obsolete by missiles around the time it deployed. While it did cost around $90 billion in 2013 dollars, it also was responsible for a good fraction of the technological things we now take for granted. Only barbarians do not remember their history, so anyone involved in modern technological projects should study it for lessons in engineering practice on long term and large scale projects.

First lesson I take from the SAGE system: solving the right problem. SAGE solved an important problem, that of air defense from enemy aircraft. It did so beautifully. The problem was, by the time it was deployed, bomber attack was a secondary issue: the primary threat was ballistic missiles. The US probably needed something like it anyway, but it is worth noticing that long-term, large-scale projects could very well be made redundant by deploy time.

Second lesson I take from the SAGE system: assemble a team that knows how to solve similar problems. MIT already had a computer on hand which constituted half of the solution, so it made perfect sense to scale up some parts of MIT into the MITRE corporation and Lincoln labs. Let’s say the government got a bug up its ass to spend $100 billion building a computer with the same capabilities as a dog’s brain, or one that could program itself. What institution would be most qualified to do this? I can’t answer this question, because pretty much nobody knows how to do something similar.

The third and final lesson I’ll take from the development of SAGE: break down the problem into manageable pieces, and solve them. They used the technology on hand in the 50s; vacuum tubes, telephone lines and CRTs. They didn’t postulate any significant breakthroughs in order to get ‘er done. They made do with what they knew was possible As such, the path to success was obvious. Engineering genius came along the way. If you don’t have manageable pieces, you don’t have a real project: you have a wish. What are the manageable pieces needed to make “nanotech” or controlled nuclear fusion a reality? What are the manageable pieces needed to make quantum computing or deriving all electrical power from the sun a reality? I don’t know, and I don’t know of anybody else who does: therefore, such things do not count as legitimate long term projects.

“One of the outstanding things… was the esprit de corps—the spirit that pervaded the operation. Everyone had a sense of purpose—a sense of doing something important. People felt the pressure and had the desire to solve the air defense problem, although there was often disagreement as to how to achieve that end. Energy was directed more toward solving individual problems, such as making a workable high-speed memory or a useable data link, than it was toward solving the problem of the value of the finished product. It was an engineer’s dream.” –John F. Jacobs Former Senior Vice President The MITRE Corporation

Insanely fascinating SAGE manuals found here:

68 Responses

Subscribe to comments with RSS.

  1. neutrino_cannon said, on March 28, 2013 at 6:18 am

    My understanding is that Soviet interceptors were also heavily dependent on ground stations to direct them towards enemies. I’m not clear on the details of how their system worked, but would be interested to know. At times, Soviet science and engineering approximates a parallel universe where the same problems were solved in different ways. Phage therapy instead of antibiotics, auxiliary lift jets on their VTOL strike aircraft (granted, the French played with those too), nuclear-powered ice breakers, and all sorts of weird solutions to the problem of powering electronics in Siberia, from the kerosene-powered radios to the rampant use of radioisotope thermoelectic generators…

    This is also why I’m a bit skeptical that stealth aircraft would have been particularly useful against the Soviet Union. If all their interceptors were being vectored to the threat by big, ground based radars, which have longer wavelengths and so aren’t fooled by stealth intended to foil short-wave fighter-mounted radar sets…

    • Scott Locklin said, on March 28, 2013 at 7:19 am

      I have a “how to shoot down a stealth fighter” rant which has been in my “to finish” pile for almost a year now. I don’t know if the Soviets could have figured it out on a large scale, but a Hungarian baker in Serbia managed it pretty well. Not sure I will ever post that one. Most of the pilots I have known are solid peckerwoods. Most of their missions are incredibly stupid, but I’d feel horribly guilty if one got killed because of me stating the obvious on the internets

      Soviet contributions to information theory and chaos theory are still ahead of the state of the art in the “free world” in some ways. A lot has been forgotten. The phage therapy thing … if it works the way it looks like it works; there is gold in them there hills. I also have something brewing on the Ekranoplan. I have some spare time; maybe I’ll finish it.

      I’d dearly love to know about the Soviet defense grid. I’m pretty sure some of it is declassified and knowable, but alas, my Russian is very rudimentary.

      • Jean-Marc Liotier said, on March 28, 2013 at 10:10 am

        > I’d dearly love to know about the Soviet defense grid

        You might start with the Vozdukh-1, which is the ground to air segment of the GCI network. It appeared in the early sixties. Fighters carried a receiver for the Lazur datalink. The Su-15’s Wikipedia page mentions SAU-58 (sistema automaticheskogo upravleniya, automatic control system) which is capable of carrying out completely automatic, ‘hands-off’ interceptions until the last moments of the interception – apparently not unlike how Sage worked with the F-106 Delta Dart (see http://www.f-106deltadart.com/sage_system.htm)

      • Bitrat said, on March 29, 2013 at 2:37 am

        It’s already been revealed – old style 50 mhz radar (think the ones used in WW II) “see” stealth craft just fine – their reflective coatings, etc are designed for microwaves…..

      • neutrino_cannon said, on March 29, 2013 at 6:00 am

        I would love to see your Ekranoplan post. Speaking of madness distilled into material form…

        • Scott Locklin said, on March 29, 2013 at 6:09 am

          This should tide you over:

  2. RealUlli said, on March 28, 2013 at 12:34 pm

    I can at least try to answer one of the questions you stated about large scale manageable pieces: Deriving all electrical power from the sun is not hard, given today’s technology. All the pieces are there, they’re manageable, they’re just incredibly expensive on that scale. Also, there are quite a few groups that don’t want it to happen, e.g. big oil and the nuclear industry.

    There are several interesting projects or experiments:

    * The Belgian rail company did a test where they roofed some track with solar cells. It turns out, if they roofed all their network, they could power all their trains with solar power. (I can’t find a link at the moment)
    * Desertec: http://en.wikipedia.org/wiki/Desertec (Large scale solar-thermal electric generation, plus transmission over long distances)

    I don’t see why projects like Desertec couldn’t be used to generate the electricity we need. I heard about similar projects being under way in China and even the US, so at least that question seems answered. 🙂

    What’s nice about the concept in Desertec, their method of using molten salt as energy transfer agent can be used to store the energy for use during the night (just keep a large tank of hot molten salt around that gets filled during the day), eliminating a lot of the storage issues other photoelectric generators face.

    • Vance Bristow said, on March 28, 2013 at 7:25 pm

      Is there a possibility that desertec could have dangers similar to the problems caused by underground disposal of industrial waste similar to that occurring in OK recently. Where I live a fault that had ben stable and no signs of stress or movement had ever been seen was caused to slip by this underground storage technique.

      • RealUlli said, on March 28, 2013 at 10:36 pm

        I think the potential environmental impact depends on the kind of salt you use. The installation itself should be mostly safe, since it’s mostly mirrors and (hopefully) no smoke. You could consider the cooling requirements an environmental impact, but I think it’s possible to even turn that into something beneficial, e.g. by using the waste heat to desalinate water (by using the low pressure steam from the steam turbines to heat salt water to near the boiling point and recondensing the resulting steam. If you do that in several stages, you end up with a lot of hot, clean water. I’m just not sure how feasible this approach is in the real world.)

        If you can think of other concerns, I’d be curious. 🙂

  3. cyberviewer said, on March 28, 2013 at 1:28 pm

    Reblogged this on Life is not a problem to be solved, its a mystery to be lived and commented:
    The Largest Computer Ever Built!

  4. Jay O'Dell said, on March 28, 2013 at 9:32 pm

    Do you have a list of the sites? I suspect some of the remnants of those buildings could still be standing.

    • Scott Locklin said, on March 28, 2013 at 9:45 pm

      The F106 website has a list, with links to photos and such. I think the buildings are mostly still there; probably too costly to tear them down. Fun fact; one of the “Planet of the Apes” movies featured some of the hardware.

      • Mike Loewen said, on March 29, 2013 at 2:02 pm


        In addition to “Conquest of the Planet of the Apes” and “Battle for the Planet of the Apes”, surplus SAGE equipment found its way into dozens of other movies and TV shows. For more than you ever wanted to know, see my AN/FSQ-7 tribute site:


        • Scott Locklin said, on March 29, 2013 at 11:12 pm

          Thanks for the link!

    • grayscones said, on March 29, 2013 at 1:30 am

      Minot AFB still has its SAGE building, last I knew it was occupied by much of the 5th Bomb Wing. See it on Google Maps at https://maps.google.com/?ll=48.413949,-101.323326&spn=0.002891,0.008256&t=h&z=18

    • TNC said, on March 31, 2013 at 11:07 pm

      The direction center in Sioux City Iowa is still standing. The building is now a food processing facility. I worked in the building for a few years and you can still see some of the features of the individual floors. I did manage to find a original piece of hardware. It appeared to be a telephone junction box about 5ft wide x 3ft tall x 2ft deep. I pulled the asset tag off of it and did a quick search of some of the numbers an IBM/Airforce job number if I recall correctly. Nearby there was 6 large conduits that was capped off and headed east of the building. I would assume that they traveled under the interstate to the ATT/MCI center that was a few miles away. I had pictures of it but lost them when my phone broke.

      This building was/is on Iowa Air National Guard property that back in the day was an F16 base and before that a Bomber school.

    • Edward Stalker said, on February 22, 2023 at 10:20 pm

      The Blockhouse at McGuire AFB was converted to office space for various tenant organizations, such as 21st AF and the various AMOG Units when I was there (1999-2003)

  5. crocodilechuck said, on March 29, 2013 at 3:21 am


    Your 3rd lesson, “They used the technology on hand in the 50s; vacuum tubes, telephone lines and CRTs” is reminiscent of Kelly Johnson’s Skunk Works, and the U-2 in particular. Rather than design a whole new airframe, he re used the fuselage of the F-104 Starfighter, and attached glider wings, to save time and money. As I recall, he gave CIA Program Manager Richard Bissell a cheque after turning over the aircraft-unspent money from the program!

    Love these posts, thanks.

    • Scott Locklin said, on March 29, 2013 at 11:11 pm

      Jerry Pournelle has some interesting thoughts on this topic as well, as his involvement in the military industrial complex came from the golden era.

      • neutrino_cannon said, on April 9, 2013 at 6:44 am

        Interesting you should mention Mr. Pournelle, as he has just linked you:


        • Scott Locklin said, on April 10, 2013 at 8:19 am

          I knew I made it as a writer when Jerry Pournelle told me I was full of shit on his blog (I disagree of course, but I have respect for my respectable elders and was tickled he noticed). His books were hugely influential on the younger me, and his thoughts on engineering and politics are classics.

          • Toddy Cat said, on April 30, 2013 at 2:52 pm

            Pournelle needs to write a memoir of his time in the defense establishment. I’d certainly buy it. I mean, I like his SF, but lots of people can write good SF, whereas Pournelle’s memories of his defense work would be irreplaceable.

  6. father ramon said, on March 29, 2013 at 12:38 pm

    “a blistering 75,000 operations per second”
    “the 386 was probably approximately equivalent”

    Um, not. The 386 was app. 2 orders of magnitude faster than that. I’d say it was closer to on par with the 8086/8088 in terms of processing speed. Fewer IPS than the 808x, but wider word to make up for it.

    • Scott Locklin said, on March 29, 2013 at 11:07 pm

      I’m guessing it was within a factor of 10 of real world 386 speed, which, combined with the 32 bit architecture is “approximately equivalent” to me. Whatever speed it clocked at, it did a *lot* of work. I’m trying to imagine an 80s PC doing anything like track all aircraft in North America and failing.

  7. Ed Hannon said, on March 29, 2013 at 2:51 pm

    Brought back some old memories. I was stationed at the North Syracuse SAGE site in NY in 1968-1969. Worked in communications – I believe that it was on the third floor. Remember being told by the computer maintenance officer that the origianl SAGE memory was a rotating drum! The core memory was an upgrade!

    • Scott Locklin said, on March 29, 2013 at 10:25 pm

      There is a bit of documentation on the drum technology in the last link listed.
      I read an old book on EDSAC when I was in grad school; the main memory was sonic echoes in a tank of mercury! 256 35 bit words!

  8. Dave Jones said, on March 29, 2013 at 3:13 pm

    Very good article, Scott, thanks. I might add just two bits of information to it: 1) the SAGE computers were built by IBM, and 2) what IBM learned ion building and running SAGE, including the programming aspects, were later incorporated into the the System/360 line of mainframes.

    • Scott Locklin said, on March 29, 2013 at 10:19 pm

      I knew IBM was a contractor, but I had no idea that 360 systems were descended from this. The J programming language has some idioms which I’m pretty sure come from 360 systems, so perhaps I am using some bits of SAGE to this day in more than the usual sense!

      • Doug Bohrer said, on April 15, 2013 at 11:03 am

        I worked on programming SAGE from 1972-1974 for the 26th Air Division at Luke AFB, AZ. SAGE had another revolutionary feature you didn’t mention, a machine portable programming language called JOVIAL. While active air defense was written in the assembler, COSEAL, programs to analyze exercise results were written in JOVIAL. The JOVIAL compiler was designed with a small machine specific kernel with most of the code based on the small kernel. Porting it to a new machine meant that you only had to rewrite the kernel. JOVIAL was so portable that when I tried the ported version on an Air Force 3rd generation machine, it had the same compiler bug that it had in SAGE. I always thought that using the dollar sign as a statement terminator was a sly joke. After all, it was written by a contractor.

        Some other interesting SAGE aspects were built-in ashtrays, how you booted the machine and SAGE’s post Air Force career. The Air Force specifications called for an ash tray for the operator console. I believe the thinking was that he could stay awake at his post longer if he was smoking rather than drinking too much coffee. I saw the ash tray. To boot the machine, you used a pre-wired plug board with the startup instructions wired into the board by hand. The board was about 16 to 18 inches square and about 3 inches thick. You slid the board into a large tilt-out drawer and literally kicked the drawer closed. It was near the floor. This might be an alternate, or additional, origin for why you “boot” your computer. By 1972, the Air Force had scrapped over half the original SAGE machines. Hollywood bought some of the pieces for props. All those blinking lights behind Lt. Uhura on the Star Trek TV show were from a SAGE console. Another show, Voyage to the Bottom of the Sea, also used blinking lights from SAGE consoles.

        • Toddy Cat said, on April 22, 2013 at 6:39 pm

          Wow, if SAGE had done nothing more than bring us classic TV shows like “Voyage to the Bottom of The Sea” it would have been more worthwhile that most modern government projects…

        • John W Cress said, on April 20, 2021 at 1:12 pm

          Boot description not exactly correct. The term comes from the Air Force practice of ‘lifting oneself by ones bootstraps’, meaning to use ones military pay to put ones self through college. An early computer used a ‘bootstrap loader’ which was a few characters punched into the first card of a deck of punched cards or even hand loaded from the bitswitches. Those few character were a branch instruction to the location of the program, or the operating system. Nowadays, of course, the bootstrap loader is in some form of ROM, but is essentially the same thing.

        • Mike Loewen said, on April 20, 2021 at 3:44 pm

          While Voyage to the Bottom of the Sea used a lot of props from SAGE, Star Trek did not. For a large list of shows with SAGE props, see http://q7.neurotica.com/Q7

  9. Ron said, on March 29, 2013 at 4:12 pm

    A little memory lane here… spent 3 years Custer AFS in Battle Creek MI in the 60’s

  10. Charles Wolfe said, on March 30, 2013 at 3:18 am

    Trained for SAGE at System Development Corp (SDC), Santa Monica, Calif (1966). But, then went thru BUIC training (the Backup Interceptor system using an Burrough’s computer, which was all transiistors – porbably the first all transistor system) and spent a couple of years at the BUICK site in Fallon, Nevada. Next time I’m in Santa Monica I’ll have to look and see if the SAGE building is still there. SDC was bought by UNIVAC (now UNYSIS). SAGE was built by IBM. At one point they figured there was no longer a need for spare parts beyond what existed. After a fire at one of the installations, IBM had to reopen its vacuum tube factory. The drums were interesting: the display for the monitors was written to the drums then picked up and sent to the monitors by other circuitry. I also heard tales from those who witnessed what happened when one of the drums broke loose — had enough weight and force to plow thru the concrete walls and go flying to the ground.

    • Vic Pulver said, on July 4, 2013 at 11:42 pm

      Sadly, the building which held the Q7 and Q32, as well as all the rest of SDC, is long gone. It is now an office complex called The Water Garden. Check out Google Maps and use the one of the old SDC addresses (e.g. 2400 Colorado Blvd).

      I worked at SDC from 1968-1978. Started on BUIC III, was a field maintenance programmer at North Truro, MA initially and then went to Fallon, NV to do the “blue-suit training” when the AF took over that function.

      • Charles Wolfe said, on July 5, 2013 at 12:20 am

        Looks like you started BUIC about the time I was transferred to the CONUS project (Southeast Asia Joint computer communications system). Left SDC near the end of 1969 when funding for CONUS was cut. Almost stayed on with a project for CAL Fire, but my Forestry background did not show up when they did their internal search for people with such a background. By the time I contacted the project manager, they were fully staffed. Who was the senior person at Fallon? Dave Pyane by any chance, he was senior at Fallon when I was there and stayed on with SDC and went thru BUIC III training, etc. Not sure where he went when the sites started opening up.

        • Vic Pulver said, on July 11, 2013 at 7:28 pm

          I don’t remember a Dave Payne. When I was in North Truro, I was part of the “central team”, which handled the initial analysis of problem reports from the field and wrote the program patches. I had very little interaction with the “standard teams”, either at Z-10 or the other BUIC III facilities. When they did the blue suit training at Fallon, which was to train the AF to do what the central team did, a group from Santa Monica set it up, and then they went to Fallon along with a bunch of us from North Truro. Each of us was assigned to a specific AF person to work with (we even had facing desks – not the easiest way to program, given this was the era of stacks of program listings, coding sheets, and punch cards.)

          We were only supposed to deal directly with another SDC person if it was for employment-related issues, like preparing for our assignments back at SDC after the project ended. Otherwise we spent the whole time with our AF counterpart. If he had to work with his team leader, we accompanied him, but mostly kept in the shadows, only taking part if necessary to correct something.

          My counterpart was the only noncom of the group – a tech sergeant. Everyone else was an officer. I guess that the programs which we had – Bomarc B prelaunch and guidance – weren’t worthy of an officer. So what if these were probably the only nuclear weapons which were specifically designed to be used (by us) on US territory.

          Even my tech sergeant didn’t find this very interesting. He insisted that after an hour or so of me teaching him missile guidance, we had to go to the rec hall where he’d teach me pool ball guidance. In retrospect, he was right.

  11. Dan Sutton said, on March 31, 2013 at 5:30 pm

    Fascinating… what a great article.

  12. eletruk said, on April 1, 2013 at 8:37 pm

    “One of the outstanding things… was the esprit de corps—the spirit that pervaded the operation. Everyone had a sense of purpose—a sense of doing something important.”
    That sense is pretty much lacking in anything to do with military contracts these days. It’s not about solving problems, it’s about getting that fat check.

  13. Steve_R said, on April 4, 2013 at 5:34 pm

    How could SAGE track all of the air traffic in North America with 32-bit, 75K IPS hardware? Easy. It didn’t run Windows! Seriously, a custom-designed, single-purpose system will always beat a stack of applciations built on a general-purpose operating system. I doubt that a UNIX system could have done it either.

    • Scott Locklin said, on April 4, 2013 at 6:06 pm

      When I started grad school, I saw an 8086 (or Z-80, I forget….) running CP/M do some amazing stuff with a CAMAC crate and BASIC.

  14. starkilleratthedarkside said, on April 5, 2013 at 8:52 pm

    Your best article to date by far, Scott. A fault-tolerant, real-time, digital, 32-bit networked computer system completed on schedule in the 50’s U.S. puts to shame all engineers and scientists I have ever met in my country, who I’m sure are unable to replicate this feat even with today’s technology.. Valuable lesson indeed. The Air Force video is GOLD, and the same I have to say about the IBM manuals. Thanks!

    • Scott Locklin said, on April 5, 2013 at 9:01 pm

      “None too learned but nobly bold
      Into the fight went our fathers of old.”

      We arguably couldn’t do it today in *my* country.

      • Toddy Cat said, on April 9, 2013 at 4:41 pm

        Well stated, and unfortunately probably true…

  15. Charles Wolfe said, on April 6, 2013 at 12:18 am

    One must remember that while the total SAGE system tracked all aircraft over the US and Canada, local SAGE installations only had to track traffic in their air space, not nationwide traffic. This is no different from the FAA system: ATCCs only track what is “local”.

    • Scott Locklin said, on April 10, 2013 at 8:22 am

      It was my understanding from my reading that each one had the capability to do it all (or as much as was left) in the event that others were knocked out.

  16. Delta101 said, on April 10, 2013 at 3:18 pm

    Dear Scott, please go ahead and write the “How to down a stealth fighter” article. The Russians already know. According to the linked video, the head of “Almaz”. The design bureau responsible for Russian AAA missiles like the S400 wrote his thesis on this topic.

    • Scott Locklin said, on April 10, 2013 at 7:14 pm

      I’m sure the Russians already have some good ideas. I’m also sure we won’t be having any wars with Russia any time soon, despite the best efforts of the idiots in Congress and the State Department. Frankly, at this point, I’d probably just ask father Putin for a job if we did.
      On the other hand, there are plenty of shitty little countries we should also leave alone, who American pilots could get shot down in. Anyway, maybe I’ll finish it one of these days.

      • Delta101 said, on April 11, 2013 at 1:49 am

        Like the Serbs? 🙂

        • Scott Locklin said, on April 11, 2013 at 2:48 am

          I’m more worried about the North Koreans, Syrians or Iranians. I think we should leave them all alone as much as is physiologically possible, but they’re shitty enough countries I’d feel bad helping them.

          If I had to guess, Hungary will be the next European country the US will be meddling with. Neocon goons are already making threatening noises. Perhaps the badass baker of Belgrade will be loaned to them; I believe he was ethnically Hungarian already.

  17. Charles Wolfe said, on April 10, 2013 at 11:13 pm

    Writing after some 40 plus years of not really thinking much about SAGE (or BUIC) does challenge the memory cells. I do have a few SDC documents left in my library, but they are all unclassified papers on various topics, none of which really jog the memory on this topic. Please excuse the few “side trips” in the following.

    Each site could take over for one or more “neighbors” if they went down for some reason. I don’t recall that a single site could manage the whole system. Given the cycle time (the time for all operations to be performed before the sequence of activities repeated ) I don’t see how one computer could have managed the whole network — I doubt the cycle time is still classified, but to be safe I won’t mention it; don’t know why I should still remember it either except whenever we made a software change the impact on timing was calculated and always critical. But, then if a single system were trying to handle the whole task, there would not likely have been much air traffic or country left to manage. Trying to squeeze a lot more geography onto the display screen would also have been a bit of a challenge; when traffic was really heavy, keeping each aircraft’s tracking symbol distinct (not obscured by others) was a bit of an “eye challenge” even on what even by today’s standards were rather large screens (see second figure in the original article).

    In response to some other comments made earlier on:
    If any of the (very) old “Time Tunnel” TV shows can be found, the “system control center” is a SAGE computer console and I think one or two work stations were used also as props.

    A few times several of us thought that a game based on SAGE (computer based or on a board) would be a worthwhile endeavor, but the more we learned the more we thought it would never sell. No one would believe any such beast actually existed or could be constructed.

    • Scott Locklin said, on April 11, 2013 at 2:51 am

      Thanks for the info! What you say makes a lot of sense. I mean, the only reason to lose one of these things on the network is if a couple of cities had been vaporized.
      A SAGE game would be really neat!

    • Vic Pulver said, on July 11, 2013 at 7:43 pm

      I didn’t work on SAGE, just BUIC III. And I remember that our cycle was closely related to a bunch of stuff, including how the displays worked. I also remember two critical limits – we had a fixed size table which contained the data related to each aircraft we were tracking, and a second fixed size table which contained data for each interceptor (manned or unmanned) we were commanding. Neither of these tables were very large. If we were controlling the battle at the end of Independence Day, I don’t think we could have controlled all the fighters; not sure we even could have tracked them all.

    • David Hoodenpyle said, on February 26, 2023 at 9:19 am

      The end result was to take in info on each aircraft such as speed, altitude, and fule if friendly and mark the unknowns. Place these targets on a rotating memory drum for access to all weapons controller display consoles. There is where decisions were made, by a human Officer and NCO. Each division included several states and each division had at least 2 BUIC sites that could take over, a lot of redundancy using minimal software programming.

      • David Hoodenpyle said, on February 26, 2023 at 9:28 am

        ….in addition, the most complicated programming was the systems program known as the “MASTER CONTROL PROGRAM” (MCP). It would test and confirm workable backup modules, take failing modules off line and replace with a back up without skipping a beat. It would switch from one main processing computer to the other which ran the program. In other words check it’s own brain for competency. Cobalt

      • Charles Wolfe said, on February 27, 2023 at 5:57 am

        I still have a few of those SDC documents also. Don’t think much about SAGE or BUIC these days….

        • Mike Loewen said, on February 27, 2023 at 7:19 am

          Charles: have you scanned your SDC documents? I would be interested in seeing them. There are a large number of SAGE documents on Bitsavers, including a pile of training docs from Keesler AFB.


          • Scott Locklin said, on March 1, 2023 at 11:02 am

            Sorry the spam bucket got you; very interesting.

            • Charles Wolfe said, on March 2, 2023 at 5:51 am

              I am not sure where they even are anymore, or what they cover. I know nothing to do with SAGE or BUIC or other classified projects.

  18. Charles Wolfe said, on April 15, 2013 at 11:32 pm

    Not only did it use the wired plug-board, just like all the IBM accounting machines, but to actually boot the computer, one manually entered the binary instructions using switches on the console to tell the computer to read the (extended) boot sequence from the plug board. The IBM 360 introduced the “floppy” drive and had the boot program on the floppy and the “read the boot program from floppy” instructions were designed into the hardware (precursor of BIOS), so one just basically pushed the “on” button rather than entering them manually. I do miss the single-step switch and memory readout displays when debugging some of the programs I’ve written on the IBM PC and “clones” — would I suppose work better in the DOS world than Windows, but still…

    • Vic Pulver said, on July 11, 2013 at 7:59 pm

      BUIC III computers used those newfangled transistor thingies, so was much smaller than SAGE. Still, The BUIC computer took up eight or so refrigerator-size cabinets. There were 2 CPU units, each in their own cabinet. Then there were 8 memory modules (4K words/module, each word was 48 bits) at two memory modules per cabinet. And then there were I think two cabinets of I/O processors. I have a vague recollection of something else called message processors.

      BUIC was started from a mag tape. You went over to a memory module and stored a tape control word by using a bunch of toggle switches. Then you went over to an I/O module and using its toggle switches, entered and executed the command to forward the tape one file. You executed that command a second time, and then entered and executed the command to read one record into memory. Finally you went over to the CPU, initialized a register to point to the memory location where you had read in the tape record, and pressed run to execute what had been read in.

      • Vic Pulver said, on July 11, 2013 at 8:45 pm

        I forgot the first step – loading the tape on the drive.

        Again, each drive was in a refrigerator-sized cabinet. These were reel-to-reel drives which used to show up in movies or tv shows whenever they had a computer room.

        The Burroughs drives worked a bit differently from the more common IBM ones. The IBM drives were loaded more or less like an audio drive – take some tape off the supply reel, run it around some pulleys and past the tape head, then wrap it around the takeup reel, holding the end of the tape against the hub by putting a finger through a hole in the flange of the reel while you turned the reel a couple of times.

        Burroughs had two key differences. There was a tape leader on the takeup reel which had a stiff piece of mylar that had been spliced on. The mylar was folded over and had a slot in it. The supply reels all had a leader as well, this one had a narrow neck and ended in a D shape – sort of like a slice through a mushroom.

        You slid the supply leader through the slot in the takeup leader, then twisted it to lock in place, then took up the slack by rotating the takeup reel. There were a couple of possible problems. If you twisted the tape the wrong way, the ends would be flat together, but you would have a half-turn in the tape. It should be noticable, but sometimes people rush. This would be bad.

        Or, if you left slack in the tape, when the takeup reel started spinning, it could suddenly jerk the tape, and you could rip the head of the supply leader, or tear the end of the takeup leader, or just rip the splice apart. And then the tape could start unreeling into the drive, or whipping around, or worse.

        The other strangeness what that there was a spring-loaded follower arm which had a roller on the end and was slipped through the flanges to ride against the tape on the takeup reel. I don’t know if this was to apply pressure as the tape wrapped on the hub, or what. The problem here was that if the takeup reel wasn’t perfectly aligned, or if one of the flanges was a bit warped, the arm could rub against the flange. And in the worst of cases, the arm rubbing against the flange could cause it to crack and fly apart.

  19. Mark Shirley said, on May 17, 2013 at 11:44 pm

    Wow, this brings back some memories! I was explaining to my son about how parts of a computer system I worked on in the Air Force became old movie props. In an effort to show him, I stumbled across this thread. I had the pleasure of working on SAGE systems at Hancock Field in Syracuse, New York, Malmstrom AFB in Great Falls, Montana and McChord AFB in Tacoma, Washington. While it is amazing what the SAGE system could accomplish with its limited (by today’s standards) resources, I’d like to point out another interesting capability designed into the SAGE computer. Because a system with that many vacuum tubes could be expected to have poor reliability, it was designed with a “marginal checking” system, an elaborate means of applying controlled “over voltage” and “under voltage” to the vacuum tube control lines. During periodic maintenance, an amplidyne system controlled by the SAGE computer could be tied into the vacuum tube control circuits. The SAGE computer would continuously loop a data pattern through selected circuits while varying the voltage levels on the tubes through the amplidyne. A circuit that was becoming weak would “drop bits” and fail the loop test. On a weekly basis, as the circuit performance declined, the diagnostics program would identify the likely fault candidates so that the computer maintenance technicians could replace the vacuum tubes before a hard failure occurred. I find it amazing that a computer of this vintage was designed to assess its own heath and the manner in which it was implemented. Thank you very much for the post!

  20. Mike Loewen said, on May 18, 2013 at 3:01 am

    Mark Shirley! Long time, no see. Drop me a note at my AN/FSQ-7 site (see earlier post). I still have a printout of the pass/fail settings for running margins on Big Mem on McChord’s Q7, among other artifacts.

  21. […] feels like a relic of the Cold War. The weapons direction console in the photo is a terminal of the SAGE (Semi-Automate Ground Environment) computer network – a continent-sized weapons system […]

  22. […] acuerdo a Scott Locklin, este fue el primer sistema que utilizó monitores CRT (tubo de rayos catódicos) y el primero en […]

  23. SAGE Anomaly Chapter One - SAGE Anomaly said, on February 22, 2022 at 12:48 am

    […] The Largest Computer Ever Built Scott Locklin https://scottlocklin.wordpress.com/2013/03/28/the-largest-computer-ever-built/  […]

  24. […] Scott Locklin, c’était le premier système qui utilisait des moniteurs CRT (tube cathodique) et le premier […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: