The Nanomechanical Computer

eldavojohn writes "The BBC is reporting on a newly proposed type of nanomechanical computer that mimics J. H. Müller & Charles Babbage's work on mechanical computational devices — just on a much smaller scale. The paper is published today in the New Journal of Physics and cites three reasons to build a computer with nanomechanical transistors over bipolar-junction or field-effect transistors: '(i) mechanical elements are more robust to electromagnetic shocks than current dynamic random access memory based purely on complimentary metal oxide semiconductor technology, (ii) the power dissipated can be orders of magnitude below CMOS, and (iii) the operating temperature of such an NMC can be an order of magnitude above that of conventional CMOS.'"

Comments

Also See the works of Neil Stephenson...

[RyanFenton]

Specifically, The Diamond Age, where such specifically mechanical nanomachines, along with artificial diamond, define the era the book takes place in. I'd say it's a charming if hyper-technical story if you haven't read it - though, things get rather unsafe for some young children in terms of strong sexuality for one prominent subplot.

Anyway, the machines aren't self-replicating, but they are fabricated in microwave-style (and larger) boxes that take an elemental 'feed' of organic compounds and data. The book has some great philosophical and social content, and breaks most of the annoying characteristics of the previous 'cyberpunk'-style writing.

Ryan Fenton

Re:Also See the works of Neil Stephenson...

[Prof.Phreak]

It's also very long... and keeps going on and on, way past the point it should've ended...and then it ends suddenly and for no apparent reason.

That was the only Stephenson book I really got bored off...

Nanotech science

[the_kanzure]

From my collection:
* Nanotechnology information [archived] [2002]
* Bibliography of nanotechnology and nanoscience [pdf] [2004]
* Brad Hein's nanotechnology website
* Ned Seeman's DNA nanotech bibliography
* MEMS/nanotech reading list
* Even more publications in nanotechnology
* sci.nano archives
* The open micro/nano-manufacturing project
* Nanotech in scifi

And if anybody has links on nanomechanical synthesis, that'd be much appreciated. IIRC, nanolithography is one of the main areas of development, along with nonlinear optics to get the required precision manufacturing.

Re:Wrong, actually

[gnalre]

Wrong, actually. The machine that was built at the end of the 20'th century was built with the precision and tolerances of the 19'th century. Deliberately, to show that it was possible. There was no doubt that the capabilities to make precision devices were available at the time of Babbage. However these devices were hand crafted, bespoke items. What Babbage machine required was precision designed machinery on an industrial scale(Note the difference engine required 25,000 parts, compare that to the average part count of a watch). That had never been attempted before and could not be achieved within any reasonable timescales using the hand crafted techniques of watchmakers and the like.

However the lesson learned allowed engineers such as Joseph_Clement(http://en.wikipedia.org/wiki/Joseph _Clement) and Joseph Whitworth(http://en.wikipedia.org/wiki/Joseph_Whit worth) to learn the importance of standardization and produce the tools to mass produce precision devices and therefore paving the way to modern industrial production.

Re:You don't need our permission

[eggnoglatte]

Going from theory to practice is the science... Actually, that would be the engineering. As in "there is no rocket science, there is physics, and then there is rocket engineering". Meanwhile, publishing the initial idea for other people (like, uh, an actual engineer) to built on is very valuable.

Oh, and your premise is wrong: building a MEMS chip of a non-trivial size pretty quickly runs in the hundreds of thousands of $, even with educational discounts. So pretty much you have to get the design ready, then ask for funding to build the thing, which is what they are presumably doing.

A tuning fork in a convential chip

[mattr]

If I interpret TFA and its references (which are more useful even just as abstracts) correctly, this is not at all the "rod logic" of Stephenson/Babbage fame. It is a single transistor, built out of two metal terminals (source and drain) and a tall, thin pillar standing between them which vibrates like a tuning fork (at 300MHz - 1GHz or so).

This pillar can be charged from the terminals and by transferring charge it can switch the current. This nano-electromechanical single electron transistor (NEMSET) was invented by other researchers, TFA mainly explores electronic properties of the NEMSET and how to put them together into circuits, create circuit elements, etc. but they didn't really do any of it yet.

Mainly it can run at high temperatures, is not as fast as ordinary transistors, but seems like it could offer multivalued logic not just binary, and as for power just about anything will do, including self-excitation, environmental vibration, etc.

So while this might be just the thing for making a laptop you can use without frying your gonads, it is not what one might think when hearing the words "nanomechanical computer".

And K. Eric Drexler wrote about it in '85

[maynard]

See Engines of Creation.

Re:Wrong, actually

[tsjaikdus]

Babbage had lots of excuses. First of all, Babbage was a very wealthy man, who didn't need government funding to begin with.

Then Babbage did build a prototype, which worked flawlessly and which he used in numerous public occasions. He even programmed it to perform 'miracles' (disturbances in continuity that up till then was only thought possible by the act of God). "Darwin saw that if apparently inexplicable discontinuities could really be the result of a system of mechanical laws laid down in advance, then here was a useful analogue of the way new species could emerge entirely through natural law."

Then one of his Difference Engines (not the small proto, but a real one) was pretty much done. It could have been finished and assembled in a short period of time if conflicts had been solved (see below). But instead, this machine was in the end melted down!

Babbage got MAJOR issues. He was effectively moving a mountain with his bare hands.

Conflict of interests. He had a fierce enemy by the name of Airy. This man, who had to decide about the funding of Babbage's engines, called the machines worthless and made fun of it. He didn't like Babbage and he was willing to make things clear.

The unit of length was different for every workshop. So Babbage could not switch manufacturer. Which proved to be disastrous as Babbage and Clement didn't get along too well. For example, Clement insisted on Babbage paying for the making of his tools (nowadays, if you can't make it, you don't get the job). Also Clement didn't wanted to return Babbage's drawings (in these times of intellectual property this is absolutely unheard of).

Frequent change of government. So, momentum was lost over and over.

Babbage wanted to make something useful. He thought he needed a processor equivalent to 166 bits today. Absurd, but how could he have known? He needed to invent everything. For example his anticipatory carriage. The modern look ahead carry. The only difference was that this carry did look ahead the full 50 decimal figures. Try that with electronic gates. You will be able to do that for maybe 4 bits, then you have to cascade and ripple your carry. Well, for Babbage, this took some time apparently. For the rest of the world it took another 100 years. Can you blame him?

And many more issues. There's a lot of factual info on Babbage if you're willing to search for it. For example stuff written by Simon Schaffer.

Re:Reversible Computing?

[marcosdumay]

You should take a look on the meaning of "reversible computing".

In short, you won't be able to reverse all operations on a debuger, but it may save some money at your light bill. Personaly, I can't see how it can be used, but I'm no expert on it, and there is a lot of buzz on that.