Is this the reason behind Microsoft’s interest in GMOs and vaccines?
By Jon Markman
Microsoft wants to use synthetic DNA strands to store the world’s data. And that could make giant data centers obsolete.
There’s no doubt about it, the world’s biggest technology firms are rushing to build data centers all over the globe. And for good reason: The Internet of Things revolution is going to produce ungodly amounts of data from sensors on our bodies and in our cars, our homes, and our offices.
Just try using the Live button on your Facebook status function; you can now broadcast HD video in real-time to the world at any time, and have the expectation the film will live forever. The scope is mind-blowing.
Even more data will come from sources like tractors, grain elevators, jet engines and cockpit computers, street lights and municipal power systems, which will all be talking to each other. All of it will need to be stored and ready for complex analytics.
In 2015, Cisco Systems estimated that some 50 billion devices would be connected to the Internet by 20generating 44 zetabytes, or 44 trillion gigabytes, of data annually.
Finding a way to efficiently store this deluge of data is going to be one of next great challenges for the tech world.
That’s why some engineers have been looking at deoxyribonucleic acid, better known as DNA.
Human DNA is remarkably dense, which makes it super-efficient at storing information. And this DNA is also unstable, which makes it easy to manipulate. Plus, it can remain readable for anywhere from 1,000 to 10,000 years – way longer than any other storage system. That’s the perfect mix of attributes for data scientists.
Each DNA strand includes nearly countless combinations of four base chemicals: adenine, cytosine, guanine and thymine. Biochemists refer to these chemicals by their first letters, A, C, G and T. Normal computer data consists of 1s and 0s. So, back in April, execs at Microsoft purchased 10 million strands of synthetic DNA from a company called Twist Bioscience.
The idea was to test how well digital content could be stored on the organic material. Microsoft first translated the 1s and 0s into a digital DNA sequence of letters. It gave that sequence to Twist and asked them to duplicate it with synthetic DNA. After Twist copied the data, it gave the organic material to Microsoft for testing. Researchers at Microsoft and the University of Washington found that all the data – about 200 megabytes of digital documents, artwork and a high-resolution music video from the band OK Go! – was intact and retrievable.
Storing digital data on DNA holds tremendous potential because it overcomes all of the shortcomings of current data storage science. It doesn’t degrade with exposure to magnetic fields or to extreme temperatures. It’s not susceptible to data loss in the event of power loss – which can happen even with state-of-the-art solid-state drive technology. It’s long-lasting. As long as the medium is kept in a relatively stable state, it can last thousands of years without degradation. In fact, recently discovered wooly mammoth DNA is sufficiently robust to resurrect an animal that has been extinct for 10,000 years.
However, DNA’s density ranks as its most impressive attraction as a storage medium. Microsoft was able to store 200MB of data on a surface about the size of the point of a pencil. Researchers estimate that all of the public data on the Internet would fit in the size of a shoebox. Maybe LeBron James’ shoebox, but you get the idea.
There is still one obstacle to overcome before DNA data storage goes mainstream – the process costs too much, even for Microsoft. Researchers are working on reducing the expense. And the cost has been falling for encoding and decoding digital data to and from synthetic DNA. Hopefully, as research continues, new processes will lead to savings.