Storage systems of the future: from film to bacteria
Long-term storage of huge amounts of data is one of the extremely relevant technologies. Humanity generates more and more information, and some of this data is especially important. Over time, paper turns into dust, magnetic storage degrades, laser discs suffer from “rotting” and “bronze” of the carrier layer.
What technologies will help humanity store important data for hundreds and thousands of years? It turns out that there are not so few of them, and most of them are not fiction at all, but projects that either can already be used or will be implemented in the near future.
Long-term data storage service on photographic film
The Norwegian company Piql and the Indian data center operator Yotta have developed and launched a service for long-term storage of information on photographic film. The new system, called Yotta Preserve, is intended primarily for companies and organizations that need large archives with information of various types – from text to images.
The data is recorded on Piql’s highly light-sensitive piqlFilm polyester film. Its advantage is its durability: the film does not degrade for several hundred years. The guaranteed data retention period is 500 years, but the manufacturer claims that the data will remain readable for up to 1000 years.
According to the authors of the project, it has two advantages:
- information is recorded on a physical medium;
- storage is realized in digital form, although, if necessary, recording of analog information is also possible.
You can store any type of content, including video, audio, handwritten documents, images, etc. Film format – 35 mm familiar to photographers. Together with the archives, illustrations and instructions for decrypting data are stored – in case they need to be accessed after several generations.
By the way, Piql is also known for the fact that the company has digitized and recorded on tape GitHub data of 21 TB. The archive is kept in a specially equipped mine on the island of Spitsbergen – the notorious “Doomsday Vault”.
Quartz media – technology from the University of Southampton team
For the first time, it became known about the creation of a working technology for recording data on a quartz medium in 2013. In the so-called “memory crystals” made of quartz, information is recorded from five different characteristics, including coordinates in three dimensions, orientation and volume. The creators of the technology have learned to control the last two parameters using the polarization and intensity of the laser beam.
In 2013, a text file of 300 kilobytes was recorded on a quartz disc. Then the data was recorded using a femtosecond laser with a wavelength of 1030 nm, pulses of 8 microjoules with a duration of 280 femtoseconds at a frequency of 200 kHz. The laser was used to burn out points arranged in layers in a crystal. The distance between them was 5 μm.
The burned-out points change the polarization of the light passing through the quartz. With the help of a specialized system that includes an optical microscope and a polarizer, the information can be read. The project is gradually developing, although not particularly quickly. Since the inception of the technology, such documents and books as the Universal Declaration of Human Rights, Newtonian Optics, Magna Carta, and the Bible have been recorded on quartz discs.
According to the developers, the technology is ideal for organizations that need to store large amounts of data unchanged for a long time. These are mainly museums, libraries, industrial enterprises. “It’s amazing even to think that we have created technology that allows us to preserve documents and other information for future generations. This technology can preserve everything we have learned, ”says Professor Peter Kazansky, head of the project.
The storage period of information on such a disk is, in principle, unlimited. The main thing is that the disc is not scratched or damaged in any other way.
Quartz glass is a technology from Microsoft and Warner Bros.
Similar technology two years ago presented two well-known companies – Microsoft and Warner Bros. A joint team of specialists has created a method of recording video on quartz glass measuring 75×75 mm and 2 mm thick. Of course, this method is not intended for distribution to home media libraries, but for long-term storage of recorded data for millions of years. Of course, in a million years there may simply be no creatures able to read this information, but theoretically, the storage period is not limited by anything.
This is already a working technology, with its help the partners recorded the original 1978 picture “Superman”. The amount of recorded information is 75.6 GB.
The technology developed by Microsoft is based on more modern developments than in the previous case. The company uses ultra-fast laser optics and machine learning to record and store on silica glass. The laser creates volumetric layers of deformation of quartz at different depths and at different angles. Machine learning helps to correctly read the recorded data by decoding it.
Warner Bros., like other film companies probably, needs reliable methods for storing media content. Films are now stored in three copies, which are distributed geographically. Two digital copies and one on film. Film is a very unreliable store of information, and quartz glass, in theory, can store data for hundreds of millions of years.
DNA strands
DNA-based data storage systems can be an outlet for humankind, which generates more and more information. Compared to all other media, DNA has a phenomenal data recording density. Another advantage is that in the case of DNA, data storage under optimal conditions does not require energy, and information can be stored for hundreds of years. After several centuries, the data can be read without problems – of course, with the appropriate technology.
The newly developed technology is based on silica capsules that store individual files. DNA tags are attached to each capsule to show what is in the file. The size of such a capsule is about 6 micrometers. Thanks to such a system, scientists have managed to learn how to extract individual images with 100% accuracy. The set of files that they created is not very large – there are only 20 of them. But if you take into account the capabilities of DNA, then such a system can be scaled up to a sextillion of files.
These 20 files were encoded into DNA fragments of about 3000 nucleotides in length, which is about 100 bytes of data. One silica capsule can hold a file up to a gigabyte in size. Once the file is wrapped, single-stranded DNA labels are placed on its surface. Multiple tags can be attached to one shell to serve as keywords. For example, “red”, “cat”, “animal”. The silica capsules labeled in this way are combined into a single data library. It is not as compact as a repository made of pure DNA, but the data is not damaged in this case.
A group of keywords – tags is used to search for files. For example, if you want to find an image of a cat, the tags are orange, cat, and domestic. To search for a tiger, only “orange” and “cat”. The search speed in such a system is still very low – something about 1 KB per second.
Another trick is that each label is associated with different colored fluorescent molecules. Therefore, during the request, any capsules with the required labels will glow in a certain color. Now there are devices that use lasers to separate objects according to the color of fluorescence, so it is technically possible to separate the necessary data.
In this case, the rest of the library will not be affected, which means that the data will not be affected. It is no longer necessary to burn a haystack for the sake of finding one needle. An additional plus in the possibility of logical search with different criteria. For example, query conditions can be complex: true for “cat”, false for “home”, true for “black”, etc.
Bacterial nanopores
This technology is even more complex than in the previous case, the basis of which is also DNA. EPFL Scientists offered use to store media, i.e. DNA strands, nanopores. They are created by EPFL bacteria in living cells using aerosilin toxin. According to scientists, nanopores can be used to record and decode information in digital form.
Reading is done using polymers – so that the process is fast enough, the speed of their passage through the nanopore is optimized. In this case, the signal of the passage of the polymer must be identified in 100% of cases in order to avoid errors in the initial data.
Machine learning is also used here – it helps to decode information from polymers, albeit so far in a very small amount. Such a system, according to the creators of the method, is much cheaper than using pure DNA strands or in the same glass capsules mentioned above.
As far as you can understand, now the technology is still too complex to be used for solving practical problems. But as a reserve for the future, this is a great option. “Our method opens up new horizons for polymer-based data storage with important benefits in the form of ultra-high density, long-term storage and device portability,” the scientists said.
Molecular storage
Another study, this time by scientists from Brown University, USA, opened up new possibilities for storing huge amounts of data in an extremely small area of the drive. Here, it is no longer DNA that is used as a carrier, but low molecular weight metabolites. Information is recorded by the interaction of metabolite molecules with each other. As a result of this interaction, new electrically neutral particles are formed, which are the elements of the recorded information.
This method is even more distant from practical implementation than nanopores, since information can be read here only in the case of chemical analysis in a specialized laboratory. But as a reserve for the future, the method is quite suitable.
