26-03-2025
The Legacy Of MIT's Whirlwind Defense System
CAMBRIDGE - MAY 20: (Pictured left to right) Narrator Charles Romine talks Jay W. Forrester, ... More Director of the Digital Computer Laboratory at MIT, in front of Whirlwind I, a large high-speed digital computer, during production of "The Search" a CBS special presentation and educational program, (Episode titled 'Robot Machines') at the Automatic Control Research Center at Massachusetts Institute of Technology (MIT). Image dated: May 20, 1954. Cambridge, Massachusetts. (Photo by CBS via Getty Images)
It was an inflection point – a turning in time. All over the world, countries were renovating their military technology. The Axis powers were rebuilding after the horror of World War II was finally done - Allied powers were also assessing their defense systems, and planning for the future.
Quietly, teams at MIT were working on any number of high-tech military projects. One of those projects would go on to become vital to national defense in America. The Whirlwind computer was a unique technology of its time, and a combination of multiple ideas about how to innovate, as the analog world was quickly becoming a digital one.
I recently got hold of a trove of resources from MIT librarians and experts on whirlwinds development, (much thanks to Debbie Douglas and Guy Fedorkow) and it made for fascinating reading. This radar defense system revolutionized how scientists approached Cold-War air defense systems.
At the request of Commander Luis de Florez of the U.S. Navy's Special Devices Center, researchers in MIT's Servomechanisms lab started work on Whirlwind in 1944, under Jay Forrester and Gordon Brown. Robert Everett would soon come in to become second in command on the project. The team took full advantage of the revolution in electronic computation pioneered by the ENIAC machine, and subsequently by John von Neumann. Based on recommendations from Dr George Valley's Air Defense Systems Engineering Council (ADSEC), the Whirlwind team shifted to major support from USAF for the air defense application. The air defense project quickly outgrew the Servomechanisms Lab, triggering the formation of Lincoln Labs and then MITRE to bring the Semi-Automatic Ground Environment (SAGE) into existence.
Meanwhile, MIT's team was competing with another contender for a set of national defense project awards.
Guy Fedorkow has recently put together a document chronicling this process. He writes:
'Whirlwind, an unproven digital machine that had many obstacles yet to overcome to achieve reliable operation, was a risky bet, and was not the only contender vying to solve USAF's air defense problem. The University of Michigan's Willow Run Laboratory, which had a large joint contract with Boeing for the development of the BOMARC air defense missile,[1]
Whirlwind was a big computer. Housed in MIT's Barta building, it was 2500 square feet, with an estimated 10,000 vacuum tubes. Imagine rows and rows of standing racks encased in metal boxes, with lots of lights, and dense clusters of displays and controls, standing as high as human beings.
As a mainframe, Whirlwind was a force to be reckoned with. Other new technology included CRT displays and a light pen or light gun for inputs.
The system was expanded to 2000 words of storage before the project eventually morphed into something called SAGE or Semi-Automatic Ground Environment, which was first made operational at McGuire Air Force Base in 1958.
When you go into the archive, looking at what we have from the Whirlwind days: you see spools of paper with punched out dots. You see pages of notes where engineers manually wrote in relevant numbers in paper coding sheets looking like spreadsheet boxes.
Though trail-blazing, Whirlwind still utilized hold-overs from the manual operational systems of its day, with paper tape, and hand-written notes. But soon those early CRT displays would be iterated into more and more sophisticated interfaces, and eventually, not too much later, the enormous mainframe started its long journey (long by today's standards) to become the desktop personal computer.
In 1959, Whirlwind was decommissioned at MIT. But it had a lot of impact on how we viewed defense, and also on the evolution of mainframe computing.
That's an important part of MIT's story, and I hope that many of those who have the opportunity will go to the archive and check out everything that happened in that very interesting time.
We're at another inflection point in time, now, with artificial intelligence changing our world. We might not be in the throes of a global conflagration that threatened the entire world order, or climbing from the rubble imposed by new airborne military systems. But we do have our work cut out for us in terms of discerning what the future looks like, and it's instructive to go back and look at history as a sort of guiding input to what we value today.
