14 thoughts on “A Single-Atom Transistor?”

  1. Um understand your distinction of the creation of something is engineering. But often in microelectronics, scientist get the science applied (yes yes applied science->engineer) right for the creation/doping a diode, or Transistor. The engineer takes those components and do something useful with it.
    Engineers maybe even help on the methods of accomplishing this. Another example is are the people who discover /create a man made unstable element are they engineers or scientist?

    I venture to guess that the researchers/developers behind this were of science training/background.

    1. Any good engineer has science training/background. It’s part of the engineering curriculum if you get a degree. People who discover things are doing science. People who design and build things are engineers. Often they’re the same people. Engineering and science are activities, not professions.

      1. Agreed that the very last thing we need right now is another “intellectual hierarchy” designed by idiots so they can condescend and preen. I’ve done both in my career (though the progress I made on the science end were null results. :/ )

  2. This is borderline philosophy debate.
    In the microelectronics field these people will refer and see themselves as scientist not as engineers, met quite a few, with titles scientist or Physicist.
    In this case “Scientist” discover a method of using a single atom to control the flow of electrons. Built a experiment to prove it. Is that really a thing? Like discovering a rounded surface has reduced friction during translational motion? Yes engineering can approve on initial discovery or use the wheel to make a thing.

  3. Can you imagine the incidence of single event upsets if you tried to use one of these in space?

    I read the article to my wife, whose first reaction was “So, is this really just a flea circus? Where the people look at the instrument data, and just believe the single-atom transistors are working because that’s what they’re told?” When I told her what a mole of single-atom transistors could mean to computing speeds (like 10E+20 times), she replied “Yeah, but there’ll be some teenager who will say ‘this thing is SO slow! It’s taking forever!'”

    It’s entertaining being married to another engineer with a sense of humor.

    1. The whole structure is 4 nm, not that far from the best present processes though when they talk about 7 nm it’s the junction that’s a few hundred atoms. It’ll be a long time, if ever, before fabrication with an atomic force microscope will scale.

      Of course, if I looked, I could probably find a transistor handbook from around 1960 that had data on every transistor then available in a fairly slim book. If you’d suggested that in a few years, we’d be building circuits with millions and now billions of transistors, they would have put you in a padded room.

      1. Well to be fair it was more like 30 years, an entire human generation, from 1960 to 1990 before we saw the first million transistor chips (Intel 80486). But the interval from 1M to 1000M (1 billion) was half again as much or about 16 years from 1990 to 2006. The problem is the curve has flattened. We did not get to 1000G (1 trillion) transistor chips by 2014 and probably won’t with current forms of technology. We’re at about 25G with Nvidia just hitting the 54G number this past Thursday with the Ampere A-100 GPU announcement.

        https://en.m.wikipedia.org/wiki/Transistor_count

        https://www.zdnet.com/article/nvidia-unveils-a100-gpus-based-on-ampere-architecture/

        1. Many years ago I was replacing 24VDC vacuum tubes in an Collins R-392 military radio with MPF102 JFETs, which was a simple upgrade that involved making fake tubes out of perf board and male RS-232 connector pins, with some resistors thrown in to provide proper coil resistance in case a particular tubes heater was used in series with another tube’s heater.

          My dad had commented that a lot of guys in his generation never made the transition to transistors because they didn’t act like tubes (the specs, bias circuits, etc were too different), but that’s because bipolars were developed first. If N-channel JFETs had come first they could have just made solid-state plastic “tubes” that were pin compatible, depending on voltage ratings.

          1. Yeah a quirk of history. I remember a college lecture where at the end of a long and expansive transistorized circuit diagram drawn on the blackboard the prof stood back and proudly announced to the class that the best thing about this circuit was that it was 6AQ5 compatible. Amidst the general chuckle my lab partner at the time leaned over and asked me what a 6AQ5 was!

            As a last gasp of the tube era, Zenith began using tubes called “compactrons” that combined the functions of two tubes into one by using multiple plates and cathodes. Talk about grid explosion! They were used to cut the tube count in “portable” (luggable) B&W TV sets. Your old man would have loved them!

          2. Love the coil resistance!! Gives the proper Buck Rodgers feel to the project! And no doubt a necessity! I NEVER saw parallel heater circuits in all the stuff I tinkered with and sometimes that 6v was put to other uses “down the line” like for pilot lamps.

          3. Well never say NEVER. There were a few TV sets that had parallel heater circuits and multiple 6-12vac taps off the transformer to keep from overloading a single tap. But unless it was the CRT on a color set they usually shared several tubes per heater circuit. I don’t *recall* ever seeing a parallel heater circuit on a radio.

  4. “Scientist” has become a political buzz word, meaning a priestly class that hands down “truth” to the polity. Which is why every new technological break through covered in the press has been done by “scientists” and why AGW is an immediate threat because 97% of “scientists” say so.

Comments are closed.