Jun 3, 2020
The development of computer technology, especially early on, is deeply tied to governmental and military research. Early computers like Colossus and the Harvard Mark Iwere a big part of the Allied war effort. The internet itself is an outgrowth of a collection of US government projects. But not all of these undertakings see the light of day. A great example of this is the not-so-subtly named Project Lightning, the NSA's attempt to create a totally new type of computer in the 1950s. And while Lightning would never lead to any public facing advances, we can see an interesting story emerge from declassified documents.
One internal report from the late 1950s summarizes Project Lightning:
"Eventually we foresee that natural limitations on speed and size will be encountered, and then the inevitable advances of our opponents will corner us, so that the duel will become one of pure wits. But while we can we must maintain our superior weapons. Project LIGHTNING was set up to preserve our advantage in speed of computation."
~Lightning, HOWARD H. CAMPAIGNE
This was in the middle of the Cold War, and the fear at the time was that the USSR would overtake the US's dominance in the field of computing. But there was an issue facing any push to advance computing, the limitations that this document cite. In this era the vacuum tube was the core component used to build computers, and while that worked it wasn't an ideal solution. These tubes ran hot, were pretty large, and didn't operate all that fast. While it would have been possible to just beef up existing computer systems you'd quickly run into diminishing returns. In order to maintain technological superiority a radical change would be needed.
The plan laid out for Project Lightning is ambitious, to say the least. In that same document the end goal was described as:
"A jet plane can go one hundred times as fast as a man can run. A computer can go ten thousand times as fast as a man can compute. LIGHTNING will go ten million times as fast."
The NSA was trying to build a supercomputer. While not exactly common, supercomputers do exist in the modern day. But in the 1950s this was unprecedented. Keep in mind that this is before the first integrated circuits, and decades before the first microprocessor. The first transistorized computer, TRADIC, was a new and untested technology. This was a strange period where the future of computing wasn't entirely clear. Project Lightning was, broadly speaking, an attempt to find the computing element of the future.
Lightning would investigate a number of contenders, but one of their early and promising leads was a device known as a cryotron. These were superconductive switches invented by Dudley Allen Buck in 1953. The first prototype cryotrons were simply a core wire wrapped in a coil of a dissimilar metal. Both metals are superconductive at low temperatures, but when a magnetic field is generated by the coil the core wire becomes resistive. The catch is that for this to work the cryotron has to be kept near absolute zero in a bath of liquid helium.
But Lightning wasn't going to be using these early cryotrons. The NSA wanted to leverage an even more futuristic technology. In the latter half of the 50s Dudley Buck was able to develop a technique for creating superconductive integrated circuits, which he called "thin film cryotrons".
"Using thin films of silicon monoxide as insulation we plate layer over layer until we have a complex assembly on a microscope slide, equivalent to a vacuum tube chassis in information-handling ability but so thin that a finger tip cannot feel its presence. This method or assembly may not only get us our 1000 megacycles, but get it for us cheap, for in mass production such techniques of assembly are much cheaper than the classical wiring and soldering"
Years before semiconductor chips existed, Buck was able to etch microscopic cryotrons on to silicon chips. A thin-film cryotron chip could switch much faster than a vacuum tube, used a scant fraction of the power, and in theory could be mass produced. The wild thing is, this was within reach. At least one prototype computer was built using thin-film cryotrons. If Project Lightning had survived we may very well have seen a massively parallel supercomputer built from superconductive circuits by the end of the 1950s.
However, Project Lightning would never reach its end goal. At least not in this iteration. The project would go on for a number of years, burning funding and resources. With the limited sources available it's hard to point to an exact failure point. It's likely that the dreams of a cryotron computer at the NSA fell by the wayside with the success of the transistor, or that Project Lightning transformed into another endeavour that has yet to see the light of day.
If you want t read more about Project Lightning, there are a number of FOIA released documents floating around. This is the one that talks the most about the cryotron side of the project.
To learn more about the cryotron in general I'd recommend checking out "The Cryotron Files" by Douglas Buck and Iain Dey. It's been an invaluable source for me personally, and it's a good read in general.
To learn more about the early development of the vacuum tube, cryotron, and early computers listen to my episode on the topic: