Microsoft is exploring new storage technologies that could shrink, for example, Facebook’s exabyte cold-storage datacenter to a cubic millimeter.
Move over tape and optical. Microsoft is working with a genetics startup to test the feasibility of creating a synthetic DNA-based storage system that can archive zettabyte-scale data.
Microsoft will buy 10 million synthetic ‘long oligonucleotides’, or DNA molecules, from US startup Twist Bioscience to explore whether they can be used for long-term mass storage, the companies announced on Wednesday.
Researchers from Microsoft and Washington University argue in a new paper that existing disk-based and optical media are not keeping up with the exponential growth in demand for data storage, much of which is infrequently accessed or ‘cold’ data.
A case in point is Facebook, which has built an entire datacenter to archive one exabyte of data stored on Blu-ray.
Microsoft reckons synthetic DNA has the potential to make Facebook’s cold-storage datacenter smaller than a match head. As it notes, DNA can store one exabyte of data per cubic millimeter.
However, that potential wouldn’t necessarily lead to Facebook doing away with its Blu-ray discs. The researchers see DNA storage as the coldest layer in a storage hierarchy, chiefly due to its slower retrieval times, which are in the tens of hours, rather than minutes for tape.
Besides storage density of existing media, the other issue is durability. As the researchers highlight, disks are generally good for three to five years, tape for up to 30 years, whereas DNA storage has been estimated to have a longevity of 2,000 years.
Twist Bioscience says it’s using a silicon-based DNA synthesis platform to generate the molecules that Microsoft is purchasing.
“DNA is a promising storage media, as it has a known shelflife of several thousand years, offers a permanent storage format and can be read for continuously decreasing costs,” Twist Bioscience CEO Emily Leproust commented.
But at this stage DNA storage is only a concept and will be years away from becoming a commercial reality, according to Doug Carmean, from Microsoft Technology and Research, who is one of the authors of the paper.
“The initial test phase with Twist demonstrated that we could encode and recover 100 percent of the digital data from synthetic DNA. We’re still years away from a commercially-viable product, but our early tests with Twist demonstrate that in the future we’ll be able to substantially increase the density and durability of data storage,” said Carmean.
Still, it’s not a bad idea to be exploring alternative media such as DNA storage given that this year the digital universe is projected to reach 17 zettabytes or 17 trillion gigabytes. Current estimates are that optical disks can store about 100GB per cubic millimeter.