> At IBM, where he worked from 1952 to 1993, Garwin was a key contributor or a facilitator on some of the most important products and breakthroughs of his era, including magnetic resonance imaging, touchscreen monitors, laser printers, and the Cooley-Tukey fast Fourier transform algorithm. > > And all that was after he did the thing for which he is most famous. At age 23 and at the behest of Edward Teller, Garwin designed the very first working hydrogen bomb...
Supposedly, at Garwin's scheming, one of the creators wasn't aware the immediate application of the algorithms they were optimizing was nuclear weapons,
> "Tukey reportedly came up with the idea during a meeting of President Kennedy's Science Advisory Committee discussing ways to detect nuclear-weapon tests in the Soviet Union by employing seismometers located outside the country. These sensors would generate seismological time series. However, analysis of this data would require fast algorithms for computing DFTs due to the number of sensors and length of time. This task was critical for the ratification of the proposed nuclear test ban so that any violations could be detected without need to visit Soviet facilities.[4][5] Another participant at that meeting, Richard Garwin of IBM, recognized the potential of the method and put Tukey in touch with Cooley. However, Garwin made sure that Cooley did not know the original purpose. Instead, Cooley was told that this was needed to determine periodicities of the spin orientations in a 3-D crystal of helium-3."
https://en.wikipedia.org/wiki/Cooley–Tukey_FFT_algorithm#His...
I also don't think it's necessary to be critical about the wording above. Perhaps "was related to nuclear weapons" reads better? But it's not exactly ambiguous, especially after reading the quoted passage.
Detection of nuclear bomb explosions is central to arms control treaties. Those are pretty much the opposite of nuclear weapon creation.
The comment I replied to seems to be taking a snarky tone toward a comment that was interesting and furthering the actual topic of the post. By highlighting that it didn't take a stance:
> one of the creators wasn't aware the immediate application of the algorithms they were optimizing was nuclear weapons
Even your reply to me seems to only allow interpreting "application ... was nuclear weapons" to mean "building" or "developing" instead of "detecting" when to me it simply means "in the domain of." And certainly you agree that detection is in the space of nuclear weapons?
The obvious reason detection technology falls under the same secrecy umbrella as weapons design is because one leads quite quickly to the other if you start to think about why it might be developed.
I think it's fascinating that Garwin was at University of Chicago at the same time as Theodore Hall studied for his masters/Phd in Physics after he left the Manhattan Project. Hall left the university in 1952 for Memorial Sloan-Kettering in New York City. At exactly that time, an intense counter espionage investigation targeting Hall had been underway for several years.
Hall provided the single most detailed document of the plutonium device to the Russians. They were both child prodigies. Hall was recruited into the Manhattan Project straight out of Harvard when he was 18.
"journalist Dave Lindorff, writing in The Nation on January 4, 2022, obtained through the Freedom of Information Act, Hall's FBI file in 2021. This 130-page file included communications between FBI Director J. Edgar Hoover to the head of the Air Force Office of Special Investigations, Gen. Joseph F. Carroll, showing that Carroll had effectively blocked Hoover's intended pursuit of Hall and Sax, probably fearing that Hall's arrest would have, in the political climate of the McCarthy Era, forced the Air Force to furlough and lose their top missile expert, Edward Hall. Carroll, a former top aide to Hoover before he became the first head of the USAF OSI, ultimately allowed Hoover's agents to question Ed Hall on June 12, 1951 (with an OSI officer monitoring the interview). Within several weeks of that session, the Air Force, which had conducted and completed its own investigation into Edward Hall's loyalty (having their own investigators question him four times), promoted him to Lt. Colonel, and later Colonel, and elevated him from assistant director to director of its missile development program. The promotions were a clear slap in the face to Hoover. Ed Hall went on to complete the development of the Minuteman missile program, and then retired."
"journalist Dave Lindorff, writing in The Nation on January 4, 2022, obtained through the Freedom of Information Act, Hall's FBI file in 2021. This 130-page file included communications between FBI Director J. Edgar Hoover to the head of the Air Force Office of Special Investigations, Gen. Joseph F. Carroll, showing that Carroll had effectively blocked Hoover's intended pursuit of [Theodore] Hall and Sax, probably fearing that [Theodore] Hall's arrest would have, in the political climate of the McCarthy Era, forced the Air Force to furlough and lose their top missile expert, Edward Hall. Carroll, a former top aide to Hoover before he became the first head of the USAF OSI, ultimately allowed Hoover's agents to question Ed Hall on June 12, 1951 (with an OSI officer monitoring the interview). Within several weeks of that session, the Air Force, which had conducted and completed its own investigation into Edward Hall's loyalty (having their own investigators question him four times), promoted him to Lt. Colonel, and later Colonel, and elevated him from assistant director to director of its missile development program. The promotions were a clear slap in the face to Hoover. Ed Hall went on to complete the development of the Minuteman missile program, and then retired."
Except... [^1] https://archive.ph/Md1YG
[^1]https://nuclearsecrecy.com/nukemap/
Were they just wrong by an order of magnitude or 2 because of previously unforseen limits, like air pressure? Or maybe 100MT is not the same as 600k Hiroshimas. Casually, the blast doesn't look like it's has a similar effect.
> All of which is to say that the idea of making hydrogen bombs in the hundreds-of-megatons yield range was hardly unusual in the late 1950s. If anything, it was tame compared to the gigaton ambitions of one of the H-bomb’s inventors. It is hard to convey the damage of a gigaton bomb, because at such yields many traditional scaling laws do not work (the bomb blows a hole in the atmosphere, essentially). However, a study from 1963 suggested that, if detonated 28 miles (45 kilometers) above the surface of the Earth, a 10,000-megaton weapon could set fires over an area 500 miles (800 kilometers) in diameter. Which is to say, an area about the size of France.
Teller’s crazy ass wanted to build a 10,000MT bomb with a 1,000MT primary..
Nuclear bomb design research since the sixties is all about making them as clean and low yield as possible. These two goals are counters to each other, but once you've mastered it your nuclear arsenal becomes less of a garage queen and a lot more useful.
It's also possible that the NYTimes is conflating facts by accident. Your demo link only accounts for single big-blast effects but the way you optimize damage with nukes is lots of smaller bombs, often in a single bomb - MIRV is one name for this. The reason is pretty simple. We approximate the explosive range as a sphere and the volume of that sphere is proportional to the cube (^1/3) of the yield.
So if you increase the yield by 10x you only increase the explosive radius by something like 2.15x. On the other hand, detonate 10 bombs side by side and you increase the radius by 10x. This not only maximizes damage, but also works to further nullify any sort of anti-missile defense. And 100MT would be well more than enough to obliterate France.
You can also kind of intuit this by thinking about 100MT means. That's 100,000,000 TONS of TNT explosive capacity. That's about 1.5 tons of TNT for each and every person in France. That's just a stupidly massive absurd amount of destruction.
By 10**.5~=3.2x
It feels somehow nice and correct that an effort to reduce a 3d problem (basically a lot of our boom goes up or down, which is not necessary in most cases) to a 2d one changes it from a gain of x^(1/3) to x^(1/2).
At one extreme there are precision attacks in silos and airfields. At the other are genocide attacks where you want to kill as many people as possible and leave as much of the land uninhabitable for as long as possible.
In the middle are precision strategic attacks, where you want to leave infrastructure and buildings so you have the option to invade and take over.
In a typical thermonuclear bomb, most of the yield is still from fission, so there are lots of fission products. The purpose of the fusion is to generate neutrons to more thoroughly fission the fissionable material.
However, it's possible to design thermonuclear bombs with greatly reduced fission fraction. The extreme example of that were the Ripple tests, which the US conducted shortly before atmospheric tests were banned. These involve a secondary without a fission "spark plug", where carefully tailored implosion cause the density/temperature at the core of the secondary to reach conditions for fusion ignition, in a way very similar to how inertial confinement fusion reactors are imagined to work. The most extreme of these has been reported to derive 99% of its yield from fusion reactions. If the neutrons from these are absorbed in something with low activation the fallout could be greatly reduced.
I don't know if that's possible, but it makes sense that it would "reduce a region the size of France to ashes." Maybe the design had a lot of stages.
As to your second sentence, pretty sure that’s what a MIRV is.
Boosting (injection of deuterium/tritium into the centre of the pit) causes a large increase in yield because fusion generates lots of high energy neutrons that go on to fast fission your (still compressed) pit.
According to one weapons designer, boosting is mainly responsible for the remarkable 100-fold increase in the efficiency of fission weapons since 1945.
Usually the yield can be selected between two or more fixed levels rather than a continuously variable input. Often this is implemented with tricks such as disabling the secondary and/or disabling boosting in the primary.
I think the key is proposed weapon, instead of practical weapon. It's a fascinating video, you should check it out!
He was an active and kind participant in many Pugwash, ISODARCO and Amaldi conferences. From 1982 on, Garwin worked with Gorbatschev’s Science Advisor, Evgeny Velikhov, and other American and Russian scientist on proposals limiting nuclear arsenals and space-based weapons.
[0] https://pugwash.org/2025/05/20/obituary-and-appreciation-for...
[1] https://en.wikipedia.org/wiki/Pugwash_Conferences_on_Science...