Imagine a large data centre of this kind: it has no semiconductor chips, it consumes no energy, its land footprint is small, and yet it stores far more information than any of today’s hyperscale data centres. Instead of semiconductors, it has hundreds of vials holding a liquid — an aqueous solution of DNA molecules.

Sounds far-fetched, like science fiction? Think again. Later this month, an interesting conference is to be held at the Sarbonne University, Paris, titled ‘Storage and computing with DNA, 2025’. If you break it down to its components — DNA storage and DNA computing — you are looking at a fascinating future of information technology, one built without semiconductors and with only factory-made DNA.

Storing data — and, by extension, computing — in the form of DNA is not a new idea. As a concept, it has been around for over half a century and, as a technology, for a dozen years. As a group of Chinese scientists note in a paper published in Synthetic and Systems Biology, the idea of using DNA for data storage was outlined independently by two scientists — Norbert Wiener of the US and Mikhail Neiman (erstwhile USSR) — almost at the same time at the release of IBM’s first hard disk in the 1960s.

Evolving technology

In 2012, a team led by George Church, a professor of genetics at Harvard Medical School and a founding core faculty member of the Wyss Institute, successfully encoded his book Regenesis: How synthetic biology will reinvent nature and ourselves into DNA. This project demonstrated the potential of DNA as a high-density, long-term data storage medium. In 2020, at the University of Texas at Austin, a research team that included scientist Stephen Jones and molecular biologist Ilya Finkelstein encoded the novel The Wonderful Wizard of Oz into DNA. They developed an advanced error-correction method for accurate retrieval of data even under harsh conditions.

As such, DNA storage is not futuristic, but an evolving technology. The problem essentially has to do with the cost of synthesising DNA — $3,500 for 1 megabyte of information, according to one estimate. However, the cost is coming down, as is the ease of synthesising DNA. People are even talking of bench-top DNA synthesisers, the size of a mini fridge.

So, are we looking at an era of ‘DNA data centres’?

Karthik Raman, Professor of Data Science and AI, Department of Biotechnology, IIT-Madras, says it is “definitely possible” although a lot more work is needed.

The proof-of-concept has been established, he says.

Microsoft, which plans to spend $80 billion to build data centres in 2025, has been working on DNA data centres for some years. “Using DNA to archive data is an attractive possibility because it is extremely dense — up to about 1 exabyte per cubic millimetre,” Microsoft said a decade ago. “While this is not practical yet due to the current state of DNA synthesis and sequencing, these technologies are improving quite rapidly with advances in the biotech industry.”

Perhaps the “not practical yet” factor still holds. While DNA synthesising is getting cheaper (sequencing is already very cheap), there are other concerns, such as lack of accuracy — there is a one per cent chance that a nucleotide combination may not represent the data correctly. And one must also factor in the energy involved in synthesising DNA, though it is likely to be far less than what current data centres consume.

DNA computing

If DNA storage is here, can ‘DNA computing’ be far behind? However, the concept of a liquid computer powered by DNA is still an emerging one. A recent article speculates on the development of a DNA-powered supercomputer capable of (say) 100 billion tasks at once.

DNA storage and computing is the ‘watch this space’ area of technology, whose implications are unimaginable.

More Like This

Published on June 1, 2025