Engineers develop a computer that operates on water droplets | Hacker News Reader

Engineers develop a computer that operates on water droplets

news.stanford.edu

59 pointsDocSavage17 comments

For those who are unfamiliar with Prakash's work... He did his PhD at MIT's Media Lab on "Microfluidic Bubble Logic", which is the earlier basis for this work. Here's his thesis (big PDF): http://cba.mit.edu/docs/theses/08.09.Prakash.pdf

How to create logical gates (XOR, AND, etc) using water: http://www.blikstein.com/paulo/projects/project_water.html

And more here: http://en.wikipedia.org/wiki/Fluidics

Personally, I prefer making logic gates with magma:

http://www.bay12forums.com/smf/index.php?topic=54046.0

Curious if anyone has any other potential applications they can share? The authors listed only a few, but the possibilities seem quite vast.

Yeah I'm really trying to think. It's tough to think of something where a computer can't just process and then have some output that changes physical matter after or even during the process. So it would need to be something that requires the object that is doing the calculation to change while calculating?

I'm trying to view it in a good way.

Radiation-resistant computation comes to mind.

If they make their fluid inert (what is a relative characteristic), there are probably several application in chemistry.

What possible applications?

Meaning, what other possible real-world uses can be conceived of?

Reminds me of the water integrator, a water based computer built in the Soviet Union in the 20s: http://en.wikipedia.org/wiki/Water_integrator

And also of MONIAC, a computer from the late 40s that modeled the UK's economy using water: http://en.wikipedia.org/wiki/MONIAC_Computer

MONIAC was the first i thought about as well.

Note that it was created by the same guy that gave economics the Phillips curve. And frankly the hydraulic computer was his better work.

Why it has gotten ignored is perhaps because it is inherently non-equilibrium and packed with feedback loops, both elements that are anathema to (neo-)classical economics.

If someone wants to have a look at a modern replacement, i would recommend Minsky.

http://www.debtdeflation.com/blogs/minsky/

I assume this is works the same way as bubble memory?

https://www.gsalmasi.com/almasiconsulting/bubbles/index.html

Video embedded in the link above was not working for me, it's on youtube:

https://www.youtube.com/watch?v=0rqPmjmQOxw

If the methodology works with arbitrary fluid droplets (not just water) there could be interesting work done from blood analysis to drug synthesis.

Maybe I missed it, but it doesn't work with water. They're using ferrofluid.

Guy Steel thought of this in the 70's: http://www.catb.org/jargon/html/A/ad-hockery.html#crunchly73... (Caution: Terrible puns) Click the link below each cartoon to read the series.

And no one has, yet, mentioned Glooper from Making Money.

Windows XP?

Reply on news.ycombinator.com