Tomorrow, April 11, will be the 108th birthday of the U.S. Navy Submarine Force. This is of interest to me because for most of my 21 years of active Naval service, I worked on submarines (affectionately referred to as "bubble boats" or "bubble machines"). I never was assigned to one so I didn't have any sea time under water (unless you count the eight hours of the dependents cruise I took my wife on) but, being assigned to a submarine tender, I worked on many of them. The training that I received for those assignments was my first real introduction to computers.
When I was in college, in 1964, the school had an IBM "something" in the Computer Room, which was in the Computer Building. As engineering students we were all required to learn computer programming, beginning with machine language, then on to FORTRAN and COBOL. This computer was not large by the standards of that day. It consisted of a console about the size of a large desk, a punched-card reader and a card sorter, each about the same size as the console and two tape drives (the 5-foot tall ones like you used to see in the movies). Off to the side were several cardpunch machines, usually referred to as "key punch machines." Although I learned the basics of computer programming, I didn't learn anything about how computers worked. They were still a mystery, or "magic," to me.
When I went in the Navy I found myself in Electronics Technician Class A School, which gave me the background for my next school, SINS School. (Naturally I volunteered for SINS School; any school with a title like that HAD to be fun! Boy, was I wrong!) SINS is an acronym for Ship's Inertial Navigation System, which was the heart of the Navigation Center on SSBNs, the Navy's ballistic missile carrying submarines. The SINS was a wonderful device, if you told SINS where it was, it could tell you where you were. The good part was that you only had to tell SINS where it was at the start of your voyage, or patrol, and it would be able to tell you where you were at any time during your 60-day patrol. Remember, this was way before GPS.
In the course of learning to operate, maintain and repair the SINS we had to learn how digital computers work. The SINS consisted of a "stable platform," with gyroscopes and accelerometers mounted on it, controlled by a digital computer. The older systems used the VERDAN computer (VERsatile Digital ANalyzer) while the newer SINS used the MARDAN computer (MARine Digital ANalyzer). It was quite a complicated electronic servomechanism. The gyros would sense a movement of the platform and send a signal to the computer, which would perform its calculations and send a signal to a motor to counteract the movement of the platform. Hence the term "stable platform." This way the platform was always level and pointed in the same direction, similar to the compass card in a marine compass only much more accurate. Since the platform was "stable," the accelerometers on it could detect the accelerations of the submarine, one for north-south movement, one for east-west movement and a third for vertical movement.
These digital computers were far different from what we have today. They were constructed completely of discrete components, that is, there were no integrated circuits, or "chips." in them; just transistors, diodes, resistors, capacitors and inductors. The "CPU" as we know it today was spread over a series of circuit boards, each one performing part of the CPU functions.
The VERDAN computer came to the Navy by way of the Air Force; in fact the whole SINS system was originally designed for an Air Force missile system that was never put in to operation (the Navajo, I think) and was based on 1955 technology. When the newer SINS, the Mark 2, Mod 3 SINS, was being designed, in order to increase the computing capacity they essentially hooked two VERDAN computers together to make the MARDAN computer.
The difference between these older computers and what we use today is like the difference between night and day. I already mentioned the fact that the "CPU" functions were spread out over several circuit boards. Today's computers use an electronic oscillator (just like the one in your quartz watch) to produce the clock frequency (the "megaHertz" you hear about). MARDAN generated its clock frequency from the spinning of its hard drive. That hard drive was called the "Main Memory" and, while I don't remember exactly how large it was, I'm sure its capacity was far less than 1 Mb! A floppy disk was completely unknown to these systems. All input was either through a punched paper tape reader (for loading the program), the control panel keyboard (scale factors and your starting position could be entered using its 15 keys), or a specially modified IBM Selectric typewriter. Output was either viewed on the control panel display (a few neon lights and a series of "Nixie" tubes, which functioned similar to a digital LED display), or printed out on the Selectric typewriter.
One nice thing about these older computers was that you could almost "see" what the data was doing. Of course, in order to troubleshoot and repair these systems "in the field," it was necessary to understand how they functioned at their lowest level. A good technician would isolate a failure down to an individual component on a specific circuit board. That doesn't happen on today's systems. But, then, today's systems are so much more reliable than their predecessors that nearly every problem is software related.
We certainly have come a long way! The VERDAN computer was about as big as a medium-sized suitcase and weighed around 40 pounds. The MARDAN was about twice as big and weighed in around 70 pounds! Put together they probably had less computing power than most of today's hand-held calculators! But, as major components of our main nuclear deterrent, they kept the peace through a very unstable time.
If you have a computer-related problem or question that you would like answered in this column, please send it to the Cedar Key Beacon by email at editor@cedarkeybeacon or PCTech@islandcity.net.