EAC-1101 In March 1958, NEC finished its first digital computer, the NEAC-1101. This machine used parametrons, invented by Eiichi Goto in 1954, and was perfected by using a single-turn transformer coupling system independently devised by NEC. This computer was designed for scientific and engineering calculations, and was Japan's first computer to use floating point operations. It was capable of decimal 7-digit floating point operations. It used 3,600 parametrons, 29 types of instructions, and had average performance of 3.5ms for addition/subtraction and 8.0ms for multiplication/division. The memory employed ferrite cores (magnetic core matrix system using the 2 ACs with different frequency), and memory capacity was 256 words (32-digit configuration). The NEAC-1101 was enhanced via improvements like expanding the memory capacity to 512 words, and was used for about 8 years for scientific and engineering calculations at NEC's research laboratory. The results from developing this computer contributed greatly to the development of subsequent parametron computers at NEC.
At 8.0ms for mult/div, that would make it right on par with a Commodore 64's BASIC math routines, even though it came 24 years earlier. That's pretty freaking impressive.
The history of computing is usually described as a nice progression from vacuum tubes to transistors and then ICs. But parametrons are only one of the many forgotten technologies that popped up along the way. Magnetic core logic such as transfluxors was used in several computers. Superconducting cryotrons were going to revolutionize computers. Tunnel-diode logic was also briefly the wave of the future. Other significant technologies were microwave logic circuits and electroluminescent logic circuits. I have to say that the 1950s came up with the best names; modern technologies just don't measure up to transfluxors, parametrons, and cryotrons.
See Digital Computer Design Fundamentals, 1962, chapter 6.
Right. So many things were tried in the 1950s. There was a lot of interest in nonlinear magnetic devices. UNIVAC had magnetic amplifiers in their "Solid State" computer (1958).
Magnetic devices have the nice property that there's no wear-out mechanism.
They're just coils and magnetic materials. They just keep going, for many decades.
The Bell System loved nonlinear magnetic devices. There were lots of them, usually potted inside little grey boxes. Here's one of many on eBay.[1]
Reportedly (English Wikipedia + other sources I forgot), the popularity of magnetic amplifiers in the US after WW2 had something to do with reverse engineering German weapon systems where such amplifiers were widely used and had received a corresponding amount of engineering work. They were great at the time unless you needed high bandwidth or light weight.
My high school computing teacher had a chemistry degree, from back around then. There was so much expectation about computers, that when the chemicals he was doing his phd on seemed to have some kind of nonlinear behaviour, his supervisor got hugely excited and made him look into whether there was a way use them to build switching devices; even though it would have been highly impractical even if they had. Basically everyone was trying ways to make switches.
The quantum flux parametron is a really fascinating design and I always wondered why no one talks about it. You can get to GHz range easily and computing will be adiabatic. Its based on Josephson-junctions so you need to provide very low temperatures. I always thought it was a more promising next gen compute technology than the current quantum computers. Especially when you get to write your own SQUIDs in a SEM. Its like two long rectangles with pads at the end+thin insulator on top+two wires for the contacts for a SQUID sandwich and suddenly you can do complex quantum circuits.
"why no one talks about [all the other alternatives]"
Expanding the question because it's the same answer for all of them: The relatively steady exponential progress of transistors murdered all the competition. Nobody had any time to think about anything else when 18 months from now the transistors would be twice as fast and cheaper at the same time. It also meant a technology had to pretty much come out of the gate already better than transistors which had been through numerous doublings. It doesn't matter if you produced something that could hypothetically double every three months for ten years if it started out a hundred times slower and a thousand times more expensive than the transistors at release day. It couldn't survive long enough in the market to get funded long enough to develop that far.
It's only now that you can really get going on some alternative... and it still kind of sucks that even so you need to produce something out of the lab that beats transistors on some relevant metric right now if you want to get anywhere. Trillions of dollars of investment into transistors is hard for any tech to overcome, even if hypothetically in 10 years it could spank transistors somehow.
I don't know much about other quantum computers, so maybe this is a common feature of them, but isn't another interesting thing about this design that it would do reversible computing, making it potentially very low in energy consumption? Or is that what "adiabatic" is a shorthand for in this context?
Meanwhile, in the US, the Univac Solid State computer[1] used similar principles in its patented[3] "Solid State Logic"[2], and was also released in 1958.
My understanding is that the magnetic amplifiers used in the V2 rocket kicked off a lot of interesting uses of magnetic cores.
The quick summary is that magnetic amplifiers started in the US in 1901, but Germany came up with much better magnetic alloys during World War II. This led to a post-war boom in mag amps, which were used in industrial control, aerospace, and computers such as the Univac Solid State. These magnetic materials also led to core memory. Transistors mostly killed off mag amps, although PC power supplies used them into the 1990s.
Magnetic amps were indeed used in the V2 in the control electronics.
But they substantially predate World War II and were once widely deployed and widely understood. The technology goes back to at least the late 19th century. It was used in many of the places where we'd use an amplifier today, like to run signals in a factory where hydraulics are not practical.
They can be quite fast - if the AC supply is well into the 10s of KHz they can reproduce audio range frequencies.
Some of the first electronic radio transmitters were based on the magnetic amplifier. The radio pioneer Reginald Fessenden developed a system using his Alexanderson alternator (a super-high-speed AC generator) which produced high power (many kilowatts) of AC at the target carrier frequency of around 30 kHz. He then fed the output of that, through a large magnetic amplifier, governing the kilowatts of generator output, with a small control current from a switch or microphone -- amplitude modulation. That was c. 1916.
During WWII, Germany used magnetic amplifiers in many different kinds of military equipment, because they had the advantage of being much more rugged and reliable than the electronic devices available at that time.
After WWII, the technology of magnetic amplifiers was one of many technologies that USA took from Germany and provided to US companies (like also the Soviet Union did with their part of Germany, but while USA took the technical documentation and samples of the products, the Soviet Union moved entire factories to Russia, piece by piece).
The US were much more focused on "kidnapping" and/or incentivizing (it was often both) the scientists and engineers while the USSR just took the dead materials. The US reaped much greater benefits.
> In 1954, Eiichi Goto invented the parametron, a logic device leveraging nonlinear parametric oscillation with two ferrite cores. Unlike the vacuum tube and early transistor circuits prevalent at the time, the parametron offered remarkable stability, requiring minimal maintenance compared to vacuum tubes with short lifetime and costing significantly less than both vacuum tubes and nascent transistors. Its simplicity and reliability made it an ideal foundation for computer design. Early applications showcased its superior fault tolerance over competing technologies, such as vacuum tubes with relatively short-lifetime, slow electromechanical relays, and unstable point-contact transistors.
I think that overstates the case for the parametron though.
Even compared to contemporary vacuum tubes they were slower. Contemporary bleeding edge vacuum tube computers could reach Mhz frequencies. Parametrons were 10-15kHz.
The supposed power benefits depended on state. At idle they were significantly more efficient but the more work they did and the faster you tried to run them the exponentially more electricity they required (and heat they generated).
Lastly because the early ones were lower frequency they required physically larger inductors. Contemporary vacuum tube computers took up similar amounts of space due to cooling constraints but parametrons were bulky in ways you couldn't fix.
Parametrons were a dead-end technology almost from the get-go due to physical scaling limits. Even the earliest use of discrete transistors blew them out of the water beyond any hope that some magical physical scaling would rescue them.
Abandoned technologies are usually abandoned for very good reasons not just because they happened to lose the race or get backed by the wrong companies. No amount of investment was ever going to help parametrons catch up with transistors - just as no amount of investment was going to help vacuum tubes catch up either.
Maybe the text should include some extraneous war crime solicitation to ensure a human wrote it. Maybe we should all do so. Death to (insert random enemy here)
Elliott 803 used mostly transistors, but it indeed also used magnetic cores for some of its logic gates.
The operating principle was different than that of the parametron, but the end effect was about the same. It used highly non-linear magnetic cores, which could not be switched by one or a few current pulses coming through wires passing through them, but enough pulses would switch the core.
This kind of logic gate, where many inputs are added and the gate switches when their sum exceeds a certain threshold, can be implemented with a wide variety of devices.
The so-called resistor-transistor logic, which was used in some early computers and there were even integrated circuits based on it, works in the same way. When enough of the input resistors are connected to high voltages, the voltage raises enough to switch on the transistor that follows the resistor network.
The logic gates with magnetic cores of Elliott 803 worked in the same way, even if they summed the magnetic fields of currents through wires and the switching threshold was determined by a non-linear magnetic core.
The neurons also use a logic of the same type, i.e. they switch on when enough synapses are excited, even if there are a lot of extra complications, as some synapses are inhibitory and the threshold for switching on is not a constant, but it is variable, depending on the history of the previous excitations.
While the first public release of Fortran was in 1956, its design started in 1954 and the GOTO statement was already included in the preliminary Fortran from 1954.
Besides the inventor of parametron, there is also Kazushige Goto, who became famous after writing in 2002 the library GotoBLAS, which was for some years the fastest library for linear algebra and which inspired all later such libraries. He later worked at Microsoft, then at Intel.
What is funny is that searching right now "Kazushige Goto" in Bing, it immediately returned a confident AI answer that "Kazushige Goto was a Japanese scientist who invented the parametron in 1954 while he was a ...".
I pity those who believe any of the AI answers that most search engines force now upon their users.
I wonder if there is any relationship between the 2 Goto, but it is unlikely, because Goto is not rare among the Japanese family names.
Goto (as in "Go to") and Goto (as in the romaji version of one of 後藤, 五藤, 五島, …) is simple two things that sound/look the same when written in ASCII w/o spaces. :)
where the first character means "behind" or "back" and the second character refers to wisteria vines.
(The first character is familiar to some of the martial arts practitioners, because it has an alternative pronunciation, "kō" instead of "go", which is used in the expression "kō-kutsu-shitsu", which is normally abbreviated to "kō-kutsu", and which means "with the back knee bent".)
The last name "Goto" in this case is actually read Gotou (ごとう), but romanization methods are their own flamewar. There's much less emphasis put on the "go" part than the "tou", so pronunciation doesn't actually sound anything like the English word "goto" at all.
Imagine the spanish word gato. Now elongate the "to", and change the ga to go, and you're close.
Very common in Japan, it seems: https://en.wikipedia.org/wiki/Got%C5%8D_(surname) . (Confusingly there is both a Mariko Goto who used to be the singer in a band named Midori and a Japanese-American violinist named Midori Goto.)
EAC-1101 In March 1958, NEC finished its first digital computer, the NEAC-1101. This machine used parametrons, invented by Eiichi Goto in 1954, and was perfected by using a single-turn transformer coupling system independently devised by NEC. This computer was designed for scientific and engineering calculations, and was Japan's first computer to use floating point operations. It was capable of decimal 7-digit floating point operations. It used 3,600 parametrons, 29 types of instructions, and had average performance of 3.5ms for addition/subtraction and 8.0ms for multiplication/division. The memory employed ferrite cores (magnetic core matrix system using the 2 ACs with different frequency), and memory capacity was 256 words (32-digit configuration). The NEAC-1101 was enhanced via improvements like expanding the memory capacity to 512 words, and was used for about 8 years for scientific and engineering calculations at NEC's research laboratory. The results from developing this computer contributed greatly to the development of subsequent parametron computers at NEC.
https://museum.ipsj.or.jp/en/computer/dawn/0017.html
At 8.0ms for mult/div, that would make it right on par with a Commodore 64's BASIC math routines, even though it came 24 years earlier. That's pretty freaking impressive.
Some years later, Goto was considered harmful :-)
Argh, beat me to it, you bastard :-).
Eiichi Goto used to joke that he was always being eliminated after Dijkstra's letter came out.
Interesting additional info, IMO this doesn’t deserve the downvotes.
The history of computing is usually described as a nice progression from vacuum tubes to transistors and then ICs. But parametrons are only one of the many forgotten technologies that popped up along the way. Magnetic core logic such as transfluxors was used in several computers. Superconducting cryotrons were going to revolutionize computers. Tunnel-diode logic was also briefly the wave of the future. Other significant technologies were microwave logic circuits and electroluminescent logic circuits. I have to say that the 1950s came up with the best names; modern technologies just don't measure up to transfluxors, parametrons, and cryotrons.
See Digital Computer Design Fundamentals, 1962, chapter 6.
Right. So many things were tried in the 1950s. There was a lot of interest in nonlinear magnetic devices. UNIVAC had magnetic amplifiers in their "Solid State" computer (1958).
Magnetic devices have the nice property that there's no wear-out mechanism. They're just coils and magnetic materials. They just keep going, for many decades. The Bell System loved nonlinear magnetic devices. There were lots of them, usually potted inside little grey boxes. Here's one of many on eBay.[1]
[1] https://www.ebay.com/itm/318704409860
Reportedly (English Wikipedia + other sources I forgot), the popularity of magnetic amplifiers in the US after WW2 had something to do with reverse engineering German weapon systems where such amplifiers were widely used and had received a corresponding amount of engineering work. They were great at the time unless you needed high bandwidth or light weight.
My high school computing teacher had a chemistry degree, from back around then. There was so much expectation about computers, that when the chemicals he was doing his phd on seemed to have some kind of nonlinear behaviour, his supervisor got hugely excited and made him look into whether there was a way use them to build switching devices; even though it would have been highly impractical even if they had. Basically everyone was trying ways to make switches.
Don't forget good old relays.
The quantum flux parametron is a really fascinating design and I always wondered why no one talks about it. You can get to GHz range easily and computing will be adiabatic. Its based on Josephson-junctions so you need to provide very low temperatures. I always thought it was a more promising next gen compute technology than the current quantum computers. Especially when you get to write your own SQUIDs in a SEM. Its like two long rectangles with pads at the end+thin insulator on top+two wires for the contacts for a SQUID sandwich and suddenly you can do complex quantum circuits.
"why no one talks about [all the other alternatives]"
Expanding the question because it's the same answer for all of them: The relatively steady exponential progress of transistors murdered all the competition. Nobody had any time to think about anything else when 18 months from now the transistors would be twice as fast and cheaper at the same time. It also meant a technology had to pretty much come out of the gate already better than transistors which had been through numerous doublings. It doesn't matter if you produced something that could hypothetically double every three months for ten years if it started out a hundred times slower and a thousand times more expensive than the transistors at release day. It couldn't survive long enough in the market to get funded long enough to develop that far.
It's only now that you can really get going on some alternative... and it still kind of sucks that even so you need to produce something out of the lab that beats transistors on some relevant metric right now if you want to get anywhere. Trillions of dollars of investment into transistors is hard for any tech to overcome, even if hypothetically in 10 years it could spank transistors somehow.
"Quantum Flux Parametron" is also the most awesomest sci-fi sounding name, straight out of Buck Rogers.
https://www.youtube.com/watch?v=RgaKjVXK0KA
"SQUID Sandwich" would be a great band name.
I don't know about other countries, but squid sandwiches are very popular in Spain: https://en.wikipedia.org/wiki/Squid_sandwich
Or back to the future.
I don't know much about other quantum computers, so maybe this is a common feature of them, but isn't another interesting thing about this design that it would do reversible computing, making it potentially very low in energy consumption? Or is that what "adiabatic" is a shorthand for in this context?
Meanwhile, in the US, the Univac Solid State computer[1] used similar principles in its patented[3] "Solid State Logic"[2], and was also released in 1958.
My understanding is that the magnetic amplifiers used in the V2 rocket kicked off a lot of interesting uses of magnetic cores.
[1] https://en.wikipedia.org/wiki/UNIVAC_Solid_State
[2] https://en.wikipedia.org/wiki/Magnetic_logic
[3] https://patents.google.com/patent/US2709798A/en
I wrote an article on the history of magnetic amplifiers: https://spectrum.ieee.org/the-vacuum-tubes-forgotten-rival
The quick summary is that magnetic amplifiers started in the US in 1901, but Germany came up with much better magnetic alloys during World War II. This led to a post-war boom in mag amps, which were used in industrial control, aerospace, and computers such as the Univac Solid State. These magnetic materials also led to core memory. Transistors mostly killed off mag amps, although PC power supplies used them into the 1990s.
Magnetic amps were indeed used in the V2 in the control electronics.
But they substantially predate World War II and were once widely deployed and widely understood. The technology goes back to at least the late 19th century. It was used in many of the places where we'd use an amplifier today, like to run signals in a factory where hydraulics are not practical.
They can be quite fast - if the AC supply is well into the 10s of KHz they can reproduce audio range frequencies.
Some of the first electronic radio transmitters were based on the magnetic amplifier. The radio pioneer Reginald Fessenden developed a system using his Alexanderson alternator (a super-high-speed AC generator) which produced high power (many kilowatts) of AC at the target carrier frequency of around 30 kHz. He then fed the output of that, through a large magnetic amplifier, governing the kilowatts of generator output, with a small control current from a switch or microphone -- amplitude modulation. That was c. 1916.
During WWII, Germany used magnetic amplifiers in many different kinds of military equipment, because they had the advantage of being much more rugged and reliable than the electronic devices available at that time.
After WWII, the technology of magnetic amplifiers was one of many technologies that USA took from Germany and provided to US companies (like also the Soviet Union did with their part of Germany, but while USA took the technical documentation and samples of the products, the Soviet Union moved entire factories to Russia, piece by piece).
The US were much more focused on "kidnapping" and/or incentivizing (it was often both) the scientists and engineers while the USSR just took the dead materials. The US reaped much greater benefits.
> In 1954, Eiichi Goto invented the parametron, a logic device leveraging nonlinear parametric oscillation with two ferrite cores. Unlike the vacuum tube and early transistor circuits prevalent at the time, the parametron offered remarkable stability, requiring minimal maintenance compared to vacuum tubes with short lifetime and costing significantly less than both vacuum tubes and nascent transistors. Its simplicity and reliability made it an ideal foundation for computer design. Early applications showcased its superior fault tolerance over competing technologies, such as vacuum tubes with relatively short-lifetime, slow electromechanical relays, and unstable point-contact transistors.
I think that overstates the case for the parametron though.
Even compared to contemporary vacuum tubes they were slower. Contemporary bleeding edge vacuum tube computers could reach Mhz frequencies. Parametrons were 10-15kHz.
The supposed power benefits depended on state. At idle they were significantly more efficient but the more work they did and the faster you tried to run them the exponentially more electricity they required (and heat they generated).
Lastly because the early ones were lower frequency they required physically larger inductors. Contemporary vacuum tube computers took up similar amounts of space due to cooling constraints but parametrons were bulky in ways you couldn't fix.
Parametrons were a dead-end technology almost from the get-go due to physical scaling limits. Even the earliest use of discrete transistors blew them out of the water beyond any hope that some magical physical scaling would rescue them.
Abandoned technologies are usually abandoned for very good reasons not just because they happened to lose the race or get backed by the wrong companies. No amount of investment was ever going to help parametrons catch up with transistors - just as no amount of investment was going to help vacuum tubes catch up either.
[flagged]
No, it’s literally a paragraph from TFA.
https://ethw.org/Milestones:Parametron,_1954#Invention_of_th...
[flagged]
AI text detector is powered by AI. QED your comment is AI summary.
Your comment is AI. My tarot reader gypsy woman flagged it with 95% confidence
AI detectors have very poor accuracy on short articles.
Maybe the text should include some extraneous war crime solicitation to ensure a human wrote it. Maybe we should all do so. Death to (insert random enemy here)
It is fine to call out AI if it is annoying.
Someone hunting for things that might be AI and polluting discussions with it is a service I did not sign up for.
Besides it is my opinion that not all AI detectors are reliable and this might currently be a big problem in some cases, especially in education.
I’ll actually respond to you because everyone else has that dog piling tone as if they haven’t seen any of the other comments.
That’s a fair point. I was probably being too bullish on this and shouldn’t have said anything.
Why?
AI reply
The Elliott 803 computer (a little later) also used magnetic core logic; https://en.wikipedia.org/wiki/Elliott_803
Not sure if the principal is similar or not. (There was one down at the Bletchley computing museum many years ago when I went down)
Elliott 803 used mostly transistors, but it indeed also used magnetic cores for some of its logic gates.
The operating principle was different than that of the parametron, but the end effect was about the same. It used highly non-linear magnetic cores, which could not be switched by one or a few current pulses coming through wires passing through them, but enough pulses would switch the core.
This kind of logic gate, where many inputs are added and the gate switches when their sum exceeds a certain threshold, can be implemented with a wide variety of devices.
The so-called resistor-transistor logic, which was used in some early computers and there were even integrated circuits based on it, works in the same way. When enough of the input resistors are connected to high voltages, the voltage raises enough to switch on the transistor that follows the resistor network.
The logic gates with magnetic cores of Elliott 803 worked in the same way, even if they summed the magnetic fields of currents through wires and the switching threshold was determined by a non-linear magnetic core.
The neurons also use a logic of the same type, i.e. they switch on when enough synapses are excited, even if there are a lot of extra complications, as some synapses are inhibitory and the threshold for switching on is not a constant, but it is variable, depending on the history of the previous excitations.
Thanks; I hadn't realised the 803 used a mix - what was the choice of which to use for a particular circuit?
As seen in Dr. Stone!
It’s stays fascinating to me how ubiquitous computation is! Seems like almost anything could “run Doom”
> vacuum tubes had relatively short lifetime
I see this a lot and I'm sure it's true but my dad had some vacuum tube HiFi setup that he'd had since college and decades later it all still worked.
I guess when there are hundreds or thousands of tubes in a computer even a fairly small chance of failure will end up happening often.
>Eiichi Goto
Is it nominative determinism if it's 2 years before the introduction of Goto (1956, Fortran apparently)
Or maybe it was named in his honour?
While the first public release of Fortran was in 1956, its design started in 1954 and the GOTO statement was already included in the preliminary Fortran from 1954.
Besides the inventor of parametron, there is also Kazushige Goto, who became famous after writing in 2002 the library GotoBLAS, which was for some years the fastest library for linear algebra and which inspired all later such libraries. He later worked at Microsoft, then at Intel.
What is funny is that searching right now "Kazushige Goto" in Bing, it immediately returned a confident AI answer that "Kazushige Goto was a Japanese scientist who invented the parametron in 1954 while he was a ...".
I pity those who believe any of the AI answers that most search engines force now upon their users.
I wonder if there is any relationship between the 2 Goto, but it is unlikely, because Goto is not rare among the Japanese family names.
Goto (as in "Go to") and Goto (as in the romaji version of one of 後藤, 五藤, 五島, …) is simple two things that sound/look the same when written in ASCII w/o spaces. :)
As a family name, it seems to be:
後藤
where the first character means "behind" or "back" and the second character refers to wisteria vines.
(The first character is familiar to some of the martial arts practitioners, because it has an alternative pronunciation, "kō" instead of "go", which is used in the expression "kō-kutsu-shitsu", which is normally abbreviated to "kō-kutsu", and which means "with the back knee bent".)
Please see: “nominative determinism”
The last name "Goto" in this case is actually read Gotou (ごとう), but romanization methods are their own flamewar. There's much less emphasis put on the "go" part than the "tou", so pronunciation doesn't actually sound anything like the English word "goto" at all.
Imagine the spanish word gato. Now elongate the "to", and change the ga to go, and you're close.
TBF, do we actually know how goto the statement was intended to be pronounced?
It's like saying Bob Gif is pronounced with a J so sounds nothing like the file format.....
>It's like saying Bob Gif is pronounced with a J so sounds nothing like the file format.....
You mean, it sounds exactly like the file format...
Yes, there is only one reasonable pronunciation of the Fortran code word which is an amalgam of the English words "go to".
GIF is an initialism, with no comparable basis.
/Please see/goto/
sheesh
pronounced `go toe`
Unlike the Fortran statement which is pronounced `jo toe`.
There's also GotoBLAS, developed by Kazushige Goto. I wonder if they have any relation or if it's a relatively common name.
Very common in Japan, it seems: https://en.wikipedia.org/wiki/Got%C5%8D_(surname) . (Confusingly there is both a Mariko Goto who used to be the singer in a band named Midori and a Japanese-American violinist named Midori Goto.)
Allegedly 35th most common according to some random website[1][2]
1: https://myoji-yurai.net/searchResult.htm?myojiKanji=%E5%BE%8...
2: https://myoji-yurai.net/prefectureRanking.htm
THIS is the type of content I read this site for. Love this article.
A link halfway down the page is causing the whole thing to have wide horizontal scroll.
Fixes it but seems overkill - why isn't white-space nowrap; ever enough?Or break-word - feel like nobody uses that one even though it has the best name for what I want to do.
Anyway
Adding my upvote to the pile