To my information management colleagues: let this percolate in your mind for a minute
19 September 2017 (Kiev, Ukraine) – Last summer I wrote a piece about how scientists had figured out how to encode data in DNA. It actually started out as a joke. As I noted, in 2011 a few bioinformaticists were at their hotel after a conference in Hamburg, Germany, talking about how could they afford to store the reams of genome sequences and other data the world was throwing at them. They were getting so frustrated by the expense and limitations of conventional computing technology that they started kidding about sci-fi alternatives. They thought “What’s to stop us using DNA to store information?”
It was one of those “lightbulb moments”. True, DNA storage would be pathetically slow compared with the microsecond timescales for reading or writing bits in a silicon memory chip. It would take hours to encode data by synthesizing DNA strings with a specific pattern of bases, and still more hours to recover that information using a sequencing machine. But the point was that with DNA, a whole human genome fits into a cell that is invisible to the naked eye. For sheer density of information storage, DNA could be orders of magnitude beyond silicon – perfect for long-term archiving.
Then … bang … last April they revealed how it was done. And they said they had gotten more inspired by IBM’s announcement in 2012 that IBM scientists had created the world’s smallest magnetic memory bit, made of just 12 atoms, a breakthrough that is transforming computing by providing the world with devices that have access to unprecedented levels of data storage. You may recall the IBM nanophysicists needed to have a little fun. In that spirit, they moved atoms by using their scanning tunneling microscope to make … a movie, which has been verified by Guinness World Records as “The World’s Smallest Stop-Motion Film”. I posted it last year but an encore. It’s short … just over a minute … and the follow-up video link tells you how they did it:
But the big endeavor has been to store light-based information as sound waves on a computer chip – something researchers compare to capturing lightning as thunder. While that might sound a little strange, this conversion is critical if we ever want to shift from our current, inefficient electronic computers, to light-based computers that move data at the speed of light.
Jackie Sheng, a nanoscientist at ETH in Zurich, walked me through the points and said:
- Light-based or photonic computers have the potential to run at least 20 times faster than your laptop, not to mention the fact that they won’t produce heat or suck up energy like existing devices.
- This is because they, in theory, would process data in the form of photons instead of electrons.
- We say in theory, because, despite companies such as IBM and Intel pursuing light-based computing, the transition is easier said than done.
- Coding information into photons is easy enough – we already do that when we send information via optical fibre.
- But finding a way for a computer chip to be able to retrieve and process information stored in photons is tough for the one thing that makes light so appealing: it’s too damn fast for existing microchips to read.
So this is why light-based information that flies across internet cables is currently converted into slow electrons. But a better alternative would be to slow down the light and convert it into sound.
And that’s exactly what researchers from the University of Sydney in Australia have now done. Quoting Birgit Stiller, the project supervisor:
“The information in our chip in acoustic form travels at a velocity five orders of magnitude slower than in the optical domain. It is like the difference between thunder and lightning. The information in our chip in acoustic form travels at a velocity five orders of magnitude slower than in the optical domain.”
That quotes comes from an article which summaries the project which you can access by clicking here.
Getting back to Jackie, he says this means that computers could have the benefits of data delivered by light – high speeds, no heat caused by electronic resistance, and no interference from electromagnetic radiation – but would also be able to slow that data down enough so that computers chips could do something useful with it. For light-based computers to become a commercial reality, photonic data on the chip needs to be slowed down so that they can be processed, routed, stored and accessed. So this is an important step forward in the field of optical information processing as this concept fulfils all requirements for current and future generation optical communication systems.
The Australia team did this by developing a memory system that accurately transfers between light and sound waves on a photonic microchip – the kind of chip that will be used in light-based computers. Nature Magazine also did a story on this and the on-line version includes an animation so you can see how it works:
First, photonic information enters the chip as a pulse of light (yellow), where it interacts with a ‘write’ pulse (blue), producing an acoustic wave that stores the data. Another pulse of light, called the ‘read’ pulse (blue), then accesses this sound data and transmits as light once more (yellow). While unimpeded light will pass through the chip in 2 to 3 nanoseconds, once stored as a sound wave, information can remain on the chip for up to 10 nanoseconds, long enough for it to be retrieved and processed.
Jackie says the fact that the team were able to convert the light into sound waves not only slowed it down, but also made data retrieval more accurate. And, unlike previous attempts, the system worked across a broad bandwidth. Building an acoustic buffer inside a chip improves your ability to control information by several orders of magnitude.
The Australia team itself noted: “Our system is not limited to a narrow bandwidth. So unlike previous systems this allows us to store and retrieve information at multiple wavelengths simultaneously, vastly increasing the efficiency of the device”.