The impending data storage crisis, and how DNA will fix it

Emerging Technologies
May 6, 2019

This article is brought to you by West Oxford Advisors.Molecular Information Storage, or MIST – also known as DNA digital storage and several other names – is using biological DNA or other molecular polymers to encode information in much the same way we use semiconductor chips, magnetic tape, and optical discs. First conceived in the 1960s by Norbert Wiener – who suggested the notion of “genetic” memory for computers – and later realized by Catherine Clelland and colleagues in 1999 by storing messages in DNA microdots on paper, MIST has gained significant traction within the last five years. This is a result of advances in DNA sequencing, recent progress in DNA synthesis, and accelerating investment by government and industry for MIST development. Last month, Microsoft and the University of Washington announced the first fully-automated MIST storage system., data generation (and by extension data storage) demand is increasing dramatically. Video content, the internet of things, and social media are a few drivers of this demand. While data generation at the margin increases, forecasts suggest that most of the world’s population will be on-line by 2030, up from 51% in 2017. In 2018, global data storage reached 33 zettabytes, up from 1 zettabyte in 2010. Triangulated base case forecasts suggest this could reach between 6-19 yottabytes by 2040.At the same time, semiconductor manufacturing is already below the trajectory of Moore’s Law. Physical limits, energy consumption, and potential raw material challenges will become more acute over the next 10-20 years in a semiconductor environment distinct from historically decreasing prices for more storage. While addressing some of the shortcomings of semiconductors, tape storage is at best a temporary, placeholder solution. The world needs a scalable storage medium with a logarithmic improvement in density relative to forecast improvements in current technologies.

MIST as a long-term solution to our data storage problem

MIST is the most practical long-term storage medium conceptually available when compared to new designs of existing mediums, new environments for computing, and alternative mediums, such as etched glass storage. The three-dimensional density of polymers, such as DNA, at the very least offers several logs improvement over today’s storage mediums. There are many exaggerated illustrations of this density, but such illustrations are directionally correct: for example, 100 million HD movies stored in DNA would take up as much space as a pencil eraser.After density, MIST benefits from the development of existing tools and technologies for reading and writing specifically in DNA. A cumulative R&D investment of $16.5B over the last 40 years into DNA sequencing and synthesis provides a solid foundation for MIST to develop. It’s worth noting that recent MIST experiments demonstrated superior storage capacity and theoretical price points compared to the first few decades of traditional computing. While significant, this cumulative investment in DNA sequencing and synthesis is less than 30% of semiconductor companies’ 2018 R&D expenditure. It should not be surprising that today’s experimental MIST solutions don’t compare to current storage mediums in terms of price, capacity, and other metrics.Yet, the runway for MIST is significant as other mediums begin to level off. There are at least eight distinct approaches to MIST currently in development. These approaches are materially different from one another. Some attempt to draft behind the requirements for healthcare applications while others attempt to develop novel approaches specifically for MIST. It is not clear which approach will best toggle the levers for MIST, yet some approaches will clearly encounter more challenges than others.Beyond storage, several efforts are exploring molecular computing. These efforts generally focus on analyzing molecules simultaneously in solution. While nascent, the potential for a combination storage and compute medium adds another dimension to MIST that could participate in markets distinct from storage.

MIST: more than just an archival data storage solution

Although much of the emphasis on MIST to date focuses on archival data storage, thought leaders describe eight distinct applications. Archival data storage is most likely the largest market for MIST, but it may be one of the last to develop. Other applications – with an emphasis on relatively small amounts of data – are more proximate and represent material opportunities to generate MIST revenue in the near-term. While estimates for the MIST market reach over $35M in five years, longer term market size estimates easily extend past a billion dollars.On the path to that size, MIST should grow incrementally through the achievement of specific technical inflection points. At one end of the spectrum, if MIST fails to make demonstrable progress over the next five years – particularly on price – then the concept could gradually fade into isolated research projects in academic institutions.The West Oxford Advisors' "Pre" Report on MIST focuses on a base case scenario of material investment over time, manifesting in a moderate pace of development. More bullish scenarios (e.g. a major investment from a technology company) could dramatically accelerate the timeline as well as the size of the MIST market.


Image source: West Oxford AdvisorsAs novel technologies emerge, it can be time-consuming and labor-intensive to establish a thorough understanding in areas that may or may not justify investment. Our hope is that this series is a starting point for considering investments (e.g. financial, R&D, time) in new, innovative spaces including synthetic biology.Drawing on interviews with more than 40 investors and stakeholders in industry, academia, and government as well as 50 institutional profiles, the report is a comprehensive overview of the 2019 MIST ecosystem. We believe this is easily the most comprehensive and thorough report available on the topic to date and should be a valuable tool for assessing any kind of investment (e.g. financial, R&D, time) in MIST over the next several years.Matt Klusas, the Managing Director of West Oxford Advisors, will be speaking about DNA data storage at SynBioBeta 2019, October 1-3 in San Francisco, register here.

Related Articles

No items found.