Scientists at the University of Michigan have created an air-powered microprocessor that is able to function without an electrical power source. It runs with just pneumatic valves and a handpump that pushes air through the system. The end result is a CPU that could eventually be used in a lab-on-a-chip device aimed at developing countries where electricity is scarce.
Minsoung Rhee and Mark Burns created the chip, which reads binary 0s and 1s as air goes in and out of the valves. The valves are controlled by changing the air pressure in a chamber below the flowing air, which closes the valve when full. Using this system, the researchers managed to create an 8-bit system of flip-flops, logic gates, and shift registers that is more mechanical than electronic.
And because pneumatic valves are already used by the same team's microfluidic systems on a chip, implementing the technology shouldn't come at a great extra cost.
[via New Scientist]
I can see the military as well as practical space uses for this type of CPU. One that does not suffer from EM damage or could even be used as a backup system in case of an electrical malfunction.
Yes, an EMP proof computer, albeit rather large at the moment, on the scale of the original transistors used in the 60s.
So this is definitely cool but the proposed practical applications are a little ridiculous at this time. At this point it's a purely academic pursuit.
You can see the clock ticking. This is an 8-bit processor that draws probably on the order of watts running at something like 1 Hz. Computing power it's like 10 orders of magnitude below a modern silicon desktop processor. Size and weight wise it's probably 6 or 7 orders of magnitude larger (just a guess).
You could shield a silicon chip in a few inches of lead and come out way ahead, computing power/unit weight or volume. And have decent EMP and radiation protection...
Sure, the air computer is immune to EMP... but it's mechanical and sensitive to mechanical shock and vibration.
I'm a little disappointed at popsci for the tone of this article. It makes it sound like an up and coming technology to watch out for. While this is true of microfluidics in general, fluid-based mechanical processors are neither new nor practical. It's much more likely that low power electronics and small generators will provide computing for areas where electricity is scarce.
This is an excellent research project that the U of Michigan should be proud of, my critique is of this article and the comments posted on it only.
Why did I write this. Boredom is a powerful motivator.
Think of this as a back-up processor, that you use while your attempting to get the primary, electric, processor back up and running. You'd only use it for highly critical systems, so it wouldn't need much in the way of processing power.
Wouldn't this air-type system be hard to maintain reliability as air is very prone to expansion by fluctuations in temperatures?