Moore's Law Strikes Again!

This article from The Guardian looks at micro- and nano-machines being researched at DARPA labs right now:

Forty years ago, Gordon Moore, the co-founder of Intel, accurately predicted that the cost of processing power would halve every two years. We have come to expect devices to get smaller, cheaper and more powerful over time. Now the revolution is spreading to other types of device. The development of mems (microelectromechanical systems) has already paved the way for "lab-on-a-chip" chemical analysis. Such breakthroughs tend to come from the military rather than industry.

"Darpa was instrumental in helping support much of the initial development of lab-on-a-chip in the early 90s," says Jon Cooper, Wolfson chair of bioengineering at the University of Glasgow. "The technologies enabled a number of US startup companies to develop miniaturised chips for faster biological analysis, giving them the necessary long-term support to grow."

Cool runnings

Now Darpa is miniaturising many new devices. Some electronics require very low temperatures, such as superconducting circuits and infra-red sensors, and the entire component is chilled by a bulky cooling system. The low-power micro-cryogenic cooler program will cool only the exact spot needed.

The key element is a "micro-machined thermal isolation structure", a tiny deep-freeze made of bismuth telluride. This cools by the thermoelectric effect when a current is applied. The micro-cooler will chill a space of about four cubic centimetres down to 200 degrees below zero, using just 0.1 watts.

Lab-on-a-chip devices already use pumps to move gas or liquid. But these pumps are not able to maintain the "hard" vacuum required for devices such as mass detectors for analysing airborne chemicals and bolometers to measure irradiation. The chip-scale vacuum micropumps program aims to produce pumps capable of producing a pressure of one millionth of an atmosphere.

Some items are for specific applications. Microsensors for imaging will deliver an infrared video camera on a chip weighing just 10g; this is specifically for uncrewed aircraft and night-vision goggles. But most of the technology will simply be made available to industry for use in future military electronics. Other programs include an atomic clock on a chip, radar on a chip, gas analysers and other sensors, radio-frequency and photonic devices. Some would have multiple uses, such as the chip-scale atomic sensors program. These tiny, high-resolution sensors can be reconfigured instantly to measure temperature, pressure, magnetic fields or other environmental factors. It's an ambitious program, but the US defence sector has a record of getting the microtechnology it needs.

I believe we already have labs on a chip that can do a whole panel of medical analysis on one sample of blood at once. All of these other gadgets will turn out to have a wide variety of civilian uses, probably more than we can imagine, once the price drops to the point where people can afford to play with them. 

We will also see progress speed up as different machines interact with each other.  For example, if researchers come up with working superconductor-based computers (which might take advantage of quantum effects), those may need to be chilled to very low temperatures.  Room-temperature superconductors are taking a while to develop.  But a chip that cools itself to -200°? That might be enough to make the whole thing work.  Other interaction effects are quite reasonable, too.

The future is coming at us faster than ever before.