In July 2016, Microsoft and University of Washington researchers successfully stored and retrieved 200 megabytes of data (including cat videos, literary classics, and a Universal Declaration of Human Rights PDF) in synthetic DNA, setting a record for DNA-based data storage density and error-free retrieval. The technique converts digital 1s and 0s into DNA base sequences (A, T, G, C), synthesizes the corresponding DNA strands, and later sequences the DNA to read the data back—leveraging biology’s 3-billion-year-old information storage system for the digital age.
Why DNA Storage?
DNA can theoretically store 215 petabytes (215 million gigabytes) of data per gram—roughly 10 million times denser than current hard drives. It lasts thousands of years in cool, dry conditions (compared to magnetic tape’s decades or hard drives’ years), and it won’t become obsolete (as long as life exists, DNA sequencing will be relevant). Humanity generates 2.5 quintillion bytes of data daily; by 2040, silicon-based storage may struggle to keep up. DNA offers a long-term archival solution for rarely accessed data like government records, cultural heritage, or scientific datasets.
The Process & Challenges
Encoding data into DNA: convert binary to base-4 DNA code → synthesize DNA strands → store in vials. Retrieval: sequence the DNA → convert back to binary. The 2016 experiment achieved 100% accuracy using error-correction algorithms (similar to QR codes). However, massive challenges remain: synthesizing DNA is slow (hours to write megabytes) and expensive ($3,500 per megabyte in 2016, down to ~$1 per megabyte by 2023), reading is also slow, and random access is difficult (you can’t easily grab just one file from a DNA archive—you sequence the whole pool).
Current Status & Future
By 2023, researchers stored increasingly complex data: entire operating systems, movies, and AI models in DNA. Catalog Technologies and Twist Bioscience launched commercial DNA storage services for archival data. However, the technology remains experimental for most use cases—fine for “write once, read never (or rarely)” scenarios like legal archives or preserving human knowledge for future civilizations, but impractical for everyday computing. The field represents a convergence of biology and information technology, imagining a future where data centers are biological rather than electronic.
Sources: Nature Biotechnology (July 2016), Microsoft Research publications, Science Magazine DNA storage coverage, IEEE Spectrum reviews