The Future of Data Storage

Speculative illustration of DNA data storage systems

What happens if DNA becomes the main way the world stores data? I used foresight methods to map where the technology is heading, built a world in 2040 around it, and made a magazine from that future.

Duration

8 Months

Methods

Speculative design, strategic foresight

What if our world’s data moved from data-centres to living DNA?

Where things stand

Exploding data, fragile media, and energy‑hungry data centers create a long‑term sustainability problem. Even as bit density improves, lifespans and materials dependency limit how much we can practically preserve.

So What?

We need durable, low-energy, high-density media that can scale across centuries, not upgrade cycles.

Projected demand vs shipped capacity

Millions of petabytes, 2019–2030. New forecast numbers reveal a widening gap that DNA media could absorb for long-term archives.

Values reconstructed from the IEEE Spectrum graphic (Gartner data) for the DNA case study. Units in millions of petabytes.

Why DNA?

  1. Density
  2. DNA holds a huge amount of data in very little space.

    How much data fits into a gram?

    A single gram of synthetic DNA can hold as much data as entire fleets of conventional media. The areas show how many devices you would need to match that single gram.

    Equivalents derived from Potomac Institute’s The Future of DNA Data Storage (2021). Assumes 1 gram of DNA ~= capacity of 42B USB sticks. This is close to DNA's theoretical capacity (about 455 exabytes per gram) and assumes older, smaller devices; the highest density demonstrated so far is about 215 petabytes per gram (Erlich & Zielinski, 2017).

  3. Longevity
  4. Properly stored DNA can remain readable for centuries.

    How long does storage last?

    Estimated lifespans for common storage media versus properly stored synthetic DNA. A symlog axis makes short and very long durations comparable in one view.

    Lifespan estimates synthesized from published research (Frontiers in Bioengineering & Biotechnology, 2022) and insights from an interview with Dr. Lee Organick. Values are illustrative.

  5. Energy
  6. DNA needs no power while it sits in storage.

    Why passive archives save energy

    Relative operational energy to keep 1 PB readable over time (normalized: HDD = 100). Passive media avoid continuous power and cooling requirements.

    Operational energy index (HDD = 100) for illustration only; excludes writes/migrations. Passive DNA scenarios adapted from Frontiers in Bioengineering & Biotechnology, 2022.

How data is stored in DNA

The workflow mirrors a reversible translation: binary data is algorithmically encoded into nucleotide sequences, strands are synthesized by DNA printers, vials of DNA are shelved as passive media, and future readers sequence the strands to decode the bits back into files. Each stage has dedicated tooling, but the payload never leaves the molecular format once written.

  1. Encode
  2. Bits are mapped to A/C/G/T sequences with error-aware codecs.

  3. Synthesize
  4. DNA foundries write the sequences into physical strands.

  5. Store
  6. Strands rest in cold, dry capsules that require no power.

  7. Sequence & Decode
  8. Readers recover the strands, sequence them, and reconstruct files.

Infographic showing the DNA data storage pipeline from encoding bits into nucleotides through synthesis, storage, and sequencing.

Conceptual pipeline illustrating how digital files become DNA and back again; adapted from speculative workshop sketches.

Process

Process diagram showing the stages of the project

What do the trends say?

PubMed mentions of DNA data storage

Annual publication counts, 1984–2024. Exact-match query executed February 2024.

Source: PubMed.gov. Query (Title/Abstract scope unless noted): ("DNA data storage" OR "DNA-based data storage" OR ("DNA" AND ("data storage" OR "digital storage"))) NOT (biobank OR "sample storage" OR preservation).

Where things are heading

Trend analysis diagram

Why is this happening?

We make more data every year, and it all has to live somewhere. That means more data centres, and data centres use a lot of energy.

Drivers of data growth

Conceptual model of data generation factors

Total global carbon emissions

Source: Climatiq

Talking to a researcher

I interviewed a DNA data storage researcher at CacheDNA several times to understand where the industry is going and how she expects DNA storage to be used. That gave me the probable future to build from.

What she told me

  • Reading and writing are slow today, but expected to speed up.
  • Sequencing is still expensive, but the price keeps falling.
  • Storing data in DNA uses less energy and material than today's storage.
  • Getting data back out of DNA is getting easier.
  • Stored well, DNA can last for centuries.
Cone of plausibility diagram

Four futures

I used Dator's Four Futures to write four scenarios: Growth, Collapse, Discipline and Transformative, each starting from the trends and the interview.

Dator's Four Futures framework diagram
Growth Scenario visualization

Growth

DNA storage keeps improving. Synthesis and sequencing get cheap enough that DNA becomes a normal way to archive data.

Collapse Scenario visualization

Collapse

DNA storage stalls on cost and technical problems while data keeps growing and data centres hit climate limits. Storage runs short and people get data quotas.

Discipline Scenario visualization

Discipline

DNA storage works, but society puts strict rules on it. It is used for healthcare, research and sustainability, and privacy comes before surveillance.

Transformative Scenario visualization

Transformative

Gene editing makes DNA storage fast and cheap. Farms store data alongside crops, DNA replaces hard drives, and people hold their own data.

The world in 2040

I built out the Transformative scenario into a full world using the Protopia framework, which aims for a future that is gradually better rather than perfect.

Transformative scenario illustration

The year 2040

DNA is the main way data is stored. Farmland in the Midwest has become data vaults, where genetically modified plants hold information. Farm towns have turned into tech towns, and schools and jobs have changed with them.

What else would change

I used an implication map to trace what the 2040 world would change, first directly and then one step further out, across technology, society, the economy, the environment, politics and values.

Implication mapping diagram

Three rough prototypes

I made three quick, rough objects from the 2040 world and presented them to a board of futures practitioners.

DNA Data Storage Capsule artifact

DNA data storage capsule

A capsule that stores humanity's collective knowledge.

DNA Data Storage Plant artifact

DNA data storage plant

Plants that store personal data and react to its "health" by wilting or blooming.

Magazine from the Future artifact

Magazine from the future

An issue of Wired from 2040, set in a world that stores its data in DNA.

The magazine

The magazine got the strongest reaction from the board, because it made the future feel believable, so I took it forward. A magazine can also hold a whole world at once: every article and ad shows another part of 2040.

Final concept visualization
Physical print of the magazine

Physical print of the magazine

Presentation of the final concept

Capstone presentation

This was my senior capstone in Industrial Design. Through the Experimental Realism (opens in new tab) mentorship program, I worked with several futures practitioners along the way.