What is DNA Computing?
The computing power of today is based on increasingly teeny tiny bits of silicon, which conducts electricity, transistors. What happens when we can't make them any smaller? And then can't make the CPU in our computers any faster. In this article, we'll dive into how DNA computing works, why it is on the mind of researchers and why Microsoft and IBM are investing large sums of cash in making the technology a reality?
Quantum Computing
Welcome to biological computing. It's no surprise that the amount of data we are producing in the world is exploding. Where do we store it though? And how do we process it? Quantum computing is one way, and of course, this show has an episode on that, but there are many hard problems to solve with quantum for one, quantum computing requires an extremely cold operating environment. So cold that atom's are at an almost complete stop. Below 1 Kelvin actually, which is -272 degrees Celsius. And that requires immense amounts of energy and space-age tech to make happen. Transistors in computers pass electrical signals, a lot of signals. The transistors in the latest CPU are getting incredibly tiny, just a few nanometers. If the transistors get any smaller though than they are currently, the electrical current flowing through the transistor easily leaks out into other components nearby or deforms the transistor due to heat and then you get messy currents, which doesn't calculate anything.
 |
| Human DNA |
DNA Computing
Innovation in computing could come to a halt, what we need is another way to keep improving the performance of computing. And one very promising approach is to use the biocompatible computing device – DNA computing. It's not about finding fossilized mosquitoes in pieces of amber. DNA computing uses DNA, biochemistry and molecular biology hardware. In fact, you already own a DNA computing device. In DNA, genetic coding is represented by four different molecules called A, T, C and G. These four bits when chained together can hold an incredible amount of data. After all the human genome is encoded in something that can be packed into a single nucleus of a cell.
First DNA Computer
DNA computing – Isn't a new concept, as such. In 1964, Russian physicist Mikhail Samoilovich Neiman "expressed original ideas and principal considerations of radical miniaturization on elements for recording, storing and retrieving of digital information to the molecular-atomic level." In other words, DNA computing. It wasn't until the 1990s that actual work started with storing data on strands of DNA. When Leonard Adelman presented the first prototype of a DNA computer, it was called the TT-100 and was basically a test tube. TT was in fact short for Test Tube. And you might be thinking by now, but Lars, what does a DNA computer look like?
Well, good question. The TT-100, might've given a clue to this. The actual computing takes place when DNA is mixed and matched in a test tube. Will you need test tubes in your laptop then? I'll get to that shortly. The data we produce every second of every hour of every day has to go somewhere ,but where? Currently a lot of it ends up in the cloud. Cloud hosted storage is super cheap and it is estimated by 2025, half of the 175 Zettabytes of data produced will be stored in the cloud. Zettabyte – a billion terabytes. Microsoft is also concerned about having to store all of these cat gifs and meme generators. So they're investing in DNA computing technology. In 2019 researchers from Microsoft and the University of Washington demonstrated the first fully automated system to store and retrieve data in manufactured DNA.
They wrote, Hello.
Not as impressive as the literary works of Shakespeare or the best pizza recipes in the world, but it was done with snippets of fabricated DNA. The end goal is to reduce warehouse sized data centers into something much smaller. Now the cool thing is that the end consumer will never know, storage will just continue to be storage. And speaking of storage, the researchers at Washington University have so far managed to store one gigabyte in DNA, and yes, they did store funny cat photos. So here are four reasons why a future, that uses DNA computing make sense.
 |
| Illustration |
It's cheap
We have DNA all around us. It is in every cell of every organic living thing. There's plenty of stock available. Producing a cheaply though is the main hurdle. Currently, as we don't use actual human DNA, rather it is produced artificially, which will be cheap.
Easy to produce.
We do it all the time. DNA wants to reproduce just like the show. We can't stop now even if we wanted to.
Scalable
Scientists estimate the DNA could hold 455 Exabytes of data in all. An Exabyte equals eight quintillion bytes or 1 billion gigabytes. Formatted in DNA every movie ever made would fit inside a volume smaller than a sugar cube. Just don't put it in your coffee.
Parallel computing
DNA can perform countless calculations in parallel where classical computing quickly reaches a limit of how many parallel computations can be made. DNA computing has almost no limit. This makes it ultra fast and incredibly powerful for scenarios like machine learning. So when can you buy your own DNA computer and fill it with the entire internet?
1
The future of computing is not about extremely powerful devices in your pocket, but rather super fast connections to the cloud and other services. DNA computers will be doing huge calculations and store enormous amounts of data, but in a fixed location. Cost has to come down first though; the current cost of storing and encoding one megabyte of data is $1 million. One area that will benefit hugely is data security. DNA based security sounds like a terrible plot in a cheap sci-fi movie, but it is real well. It will be DNA cryptography works largely like classic geography with their private and public key, but because it is incredibly fast, the keys can be massive like that big.
DNA computing has the promise of cheap, huge accessible data storage and an exponential increase in computing power and speed. There are huge challenges still though, not least the cost of creating the DNA, but we are past the proof of concept phase and real money is being spent to create real commercial solutions. Do you have an opinion on DNA computing and what it could be useful.
Follow us @ Facebook : Advanced Tech World
