Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Technology Hardware

A Computer That Operates On Water Droplets 67

Posted by Soulskill from the what-voyager's-computer-is-made-of dept.
Okian Warrior notes a Stanford project to build a basic computer that operates on water droplets. One of its creators, Manu Prakash, says the goal is not to compete with digital computers for manipulating data (though they can theoretically perform all of the same computations). Instead, "Our goal is to build a completely new class of computers that can precisely control and manipulate physical matter. Imagine if when you run a set of computations that not only information is processed but physical matter is algorithmically manipulated as well." The biggest obstacle in creating the water computer was figuring out a way to develop a clock mechanism. The team decided to use a rotating magnetic field, which is both precise and easy to control. To get it to interact with the water, they put arrays of tiny iron bars on glass slides, and then added a layer of oil, and finally another glass slide. Magnetized water droplets are injected into this scaffolding, and the magnetic field can then easily push them along paths created by the iron. "It's about manipulating matter faster," Prakash said.

This discussion has been archived. No new comments can be posted.

A Computer That Operates On Water Droplets More

A Computer That Operates On Water Droplets

Comments Filter:

  • Block-Transfer Computations (Score:3)

    by Feneric ( 765069 ) on Tuesday June 09, 2015 @07:30PM (#49880159) Homepage
    It sounds almost like an early form of block-transfer computations as described in Doctor Who.
  • or something like that. i don't know.
  • Surely the CA state water resources control board will be paying them a visit soon...

  • Get the mop (Score:5, Funny)

    by PPH ( 736903 ) on Tuesday June 09, 2015 @07:52PM (#49880249)

    We have a memory leak.

  • Microfluidics? (Score:5, Interesting)

    by smaddox ( 928261 ) on Tuesday June 09, 2015 @07:55PM (#49880259)

    The project looks interesting from an academic perspective, but the stated application to biological microfluidics seems ridiculous when it requires the droplets to be filled with magnetic materials that could potentially compromise any test you might want to perform.

    Microfluidic channels are fairly easy to produce using traditional lithography, and a simple water pump produces all of the motion necessary. It's difficult to see how this really improves upon that model.

    • Re:Microfluidics? (Score:4, Informative)

      by Okian Warrior ( 537106 ) on Tuesday June 09, 2015 @09:42PM (#49880701) Homepage Journal

      Microfluidic channels are fairly easy to produce using traditional lithography, and a simple water pump produces all of the motion necessary. It's difficult to see how this really improves upon that model.

      You have a valid point, but I thought it was an interesting approach(*).

      In his paper [arxiv.org], Dr. PraKash notes that microfluidics requires pumps, valves, and other controlling hardware to route the chemicals to the required places.

      His system moves microsamples around using magnetic fields, eliminating the need for pumps and valves.

      Check out his dancing droplets [youtube.com] video on YouTube. There's really a lot going on at the atomic level with these micro droplets.

      (*) I submitted the article

    • Re:Microfluidics? (Score:4, Interesting)

      by mlts ( 1038732 ) on Tuesday June 09, 2015 @10:36PM (#49880847)

      Isn't this something similar to a switching apparatus that Russia was working on, except with switching done by jets of water, where two jets would cancel each other out, creating a zero? I remember reading about half-adders done this way, as well as far more complex hydrofluidic building blocks.

      One advantage of this form of switching is that it is EMP-proof (barring a blast that actually causes physical damage), something that has been the Achilles heel of almost all technology made today.

  • They missed the memo (Score:3)

    by drinkypoo ( 153816 ) <drink@hyperlogos.org> on Tuesday June 09, 2015 @08:08PM (#49880321) Homepage Journal

    Hydraulic, pneumatic, hydropneumatic, mechanical... these kinds of control systems, as it turns out, are less reliable than using electronics, and they react more slowly. They also don't learn. Before you get enough complexity to have them learn, see points A and B.

  • On a hot day it would give totally new meaning to the term "vaporware".

  • Cloud computing (Score:3)

    by jd2112 ( 1535857 ) on Tuesday June 09, 2015 @08:26PM (#49880401)
    Taken to a new level

  • Fluidics was very big some 25 years ago (Score:5, Interesting)

    by 140Mandak262Jamuna ( 970587 ) on Tuesday June 09, 2015 @08:30PM (#49880423) Journal
    Originally fluidics circuits [wikipedia.org] mimicked electronic circuits but all those valves controlling fluid flow were quite cumbersome. They built standard "cells" that would have a large inlet and out let for main "flow" and then additional taps which were controlled by smaller streams of the fluid. There were cells that would do the plumbing of making tap of one cell be the main inflow for another cell etc. They had the equivalents of transistor, one small bias flow controlling a large main flow to either "out/ground" or to tap A etc. They could stack them up to avoid running tubes. They could do what the primitive digital electronic circuits could do. They had built simple ripple counters, and flip flops etc. They are digital circuits, but with "power"! They had enough power to move small actuators. Enough to move baffle doors of air ducts etc. (Quoting from memory of some IEEE spectrum article read some 25 years ago. sorry for the mistakes)

    It produced some PhDs and some R&D grants. But never went beyond academic labs. I don't think there was the Big Digital Electronic conspiracy to stymie the upstart competitor. Even their proponents did not really believe it could take on digital electronic circuits.

    • It produced some PhDs and some R&D grants. But never went beyond academic labs.

      The typical automatic transmission contains a fluidic computer, and that was the only computer they ran on until the 1980s. However, today the trend is towards shift-by-wire, with no linkage present even to handle limp-home mode shifting. If the electrical system serving the trans goes south, it's dead, Jim.

      • It produced some PhDs and some R&D grants. But never went beyond academic labs.

        The typical automatic transmission contains a fluidic computer, and that was the only computer they ran on until the 1980s. However, today the trend is towards shift-by-wire, with no linkage present even to handle limp-home mode shifting. If the electrical system serving the trans goes south, it's dead, Jim.

        All the more reason to continue driving manual. Unfortunately manual transmission cars are getting increasingly hard to find as more people who don't know how to drive prefer everything automatic.

        • Re: (Score:2)

          by Ramze ( 640788 )

          Or you could just get an electric car which typically has no transmission because with its higher torque, it can go from 0 to 60 in less than 5 seconds essentially on one gear.

          I hope to have one electric car one day for everyday use and trips of less than 200 to 250 miles and another vehicle (perhaps just borrowing a parent's jeep from time to time) for longer road trips. I'm hoping for a breakthrough in battery or fuel cell tech to take things to the next level as well as a price drop - but, since Telsas

          • Motorcycle

            Porche performance, Prius gas milage.

            • Don't forget to fill out that Donor Card [telegraph.co.uk]

            • Awesome. I was just wondering where my next liver would come from.

              Don't forget Hummer emissions.

              Also, I have done lost count of the time I've been stuck behind a superbike on twisty roads. Motorcycles are slow if they're not on a prepared track that's been carefully swept of detritus. In the real world, keeping them on the road is a challenge, and you have to stick your head into oncoming traffic to take left-hand corners rapidly. I've whipped 750s and bigger like cream with stock-engine Nissan 240SX and Su

              • I don't know about the bikes in your area, but my BMW has a cat converter just like any car, and the emissions are just as clean.

                I get 55 to 65 mph off the freeways, 45 to 50 on the freeways.

                Gravel can be an issue, mainly because drivers drag it from the shoulder onto the road because they can't stay on the road. Having a dual sport with ABS, ASC and tires made for road and off road means I can generally ignore gravel that would send a streetbike spinning.

                As for being able to beat them on the road... it is

                • don't know about the bikes in your area, but my BMW has a cat converter just like any car, and the emissions are just as clean.

                  Most of the bikes on the road are still pre-emissions, because motorcycles tend to be kept longer than cars... cheaper maintenance, cost almost as much as a car but do much less.

                  As for being able to beat them on the road... it is not a race. Slow down. YMMV

                  Haha, tell that to motorcyclists.

        • All the more reason to continue driving manual. Unfortunately manual transmission cars are getting increasingly hard to find as more people who don't know how to drive prefer everything automatic.

          Here in Finland, manual transmission is the norm, presumably due to fuel efficiency, which suits our high fuel prices. If our white trash can learn to drive them, then it shouldn't be too hard for you guys either. Automatic transmissions are generally used by people who lost a leg so can't easily operate the clutch.

          • Here in Finland, manual transmission is the norm, presumably due to fuel efficiency, which suits our high fuel prices.

            Sorry, modern automatics have higher efficiency than driving stick. The only cars in which the stick gives better efficiency are cars with old-school slushboxes like 5 and 6 speeds. The 8+ speed automatics have multiple overdrives and can get better mileage than you can.

            The new automatics might still be annoying to drive, of course. For example the ZF9 has to make a weird shift to get in and out of overdrive, best-case it is much slower than the other shifts, though pretty fast. Still, as a driver what you

        • If the transmission in my Audi A8 should go tits-up, I will probably drop the two grand for a six speed swap kit from another vehicle with a 4.2 V8 in it. Allroad, S4... apparently you can swap an OD from a TDI gearbox and get really bitchin' mileage out of it, too.

          Problem is, lots of cars aren't even offered with a stick. Both my sedans fall into that category. There was no manual A8 (and only some S8s) and the only manual W126 was the AMG... and even that was an option.

      • Thanks. I did not know that. I imagined the automatic transmission to be using analog computation, along the lines of Watt's governor of steam engines. Was it digital?

        • I imagined the automatic transmission to be using analog computation, along the lines of Watt's governor of steam engines. Was it digital?

          Ultimately what it does is control various valves which can either be open or closed, so I guess you'd say so, but I'm not a transmission expert.

  • Imagine if when you run a set of computations that not only information is processed but physical matter is algorithmically manipulated as well.

    And here I thought the movement of electrons in normal computers was already the embodiment of algorithmic manipulation of physical matter. Silly me.

    • And here I thought the movement of electrons in normal computers was already the embodiment of algorithmic manipulation of physical matter. Silly me.

      Look, a brother's got to write his PhD thesis on something, you know? Don't be a spoil-sport.

  • I can't wait to see the SQRT function.

  • old tech (Score:4, Insightful)

    by NostalgiaForInfinity ( 4001831 ) on Tuesday June 09, 2015 @11:33PM (#49881011)

    A droplet computer is certainly new and interesting, but people should remember that computers using pneumatic or fluidic elements are actually quite old.

    http://en.wikipedia.org/wiki/F... [wikipedia.org]

  • Imagine.. (Score:1)

    by Anonymous Coward

    Imagine if when you run a set of computations that not only information is processed but physical matter is algorithmically manipulated as well.

    Imagine if your TI-89 could be not only a graphing calculator but a handheld Rube Goldberg machine as well.

  • The Eric S Raymonds jargon.txt or "The New Hackers Dictionary" has a series of illuminating illustrations on the features of a water-powered computer, made by Bells & Whistles incorporated.
    No cooling problems, good floating point performance, but the overflow error and subsequent core dump is to be taken seriously ...

    See http://www.catb.org/jargon/htm... [catb.org]

  • The biggest obstacle in creating the water computer was figuring out a way to develop a clock mechanism.

    Water Clock [wikipedia.org]

  • ... hum.. that's all I've got to say about that.

Slashdot Top Deals

If all else fails, lower your standards.

Close