Scientists Develop DNA-Based Memory Device Using 100x Less Power

DNA just sits there, holding information without any power at all
The fundamental advantage of DNA storage over silicon-based systems is its chemical stability and lack of power requirements.
Mark

Why does DNA use so much less power than silicon memory?

Mimi

Because DNA is chemically stable on its own. Silicon memory needs constant electricity to hold its state. DNA just sits there, holding information in its molecular structure without any power at all.

Mark

So you're saying we could store data and just... leave it alone?

Mimi

Exactly. No cooling, no constant current, no degradation from electrical stress. You encode the information once, and it stays there until you need to read it.

Mark

That sounds almost too good to be true. What's the catch?

Mimi

Speed. Reading DNA storage takes time—minutes or hours, depending on what you're looking for. It's not like pulling something off a hard drive. And writing data into DNA is also slow and expensive right now.

Mark

So it's not for everyday computing.

Mimi

Not yet. But for data that doesn't need constant access—archives, backups, long-term preservation—it could be transformative. A library storing centuries of records, or a company archiving old data.

Mark

How close are we to actually using this in real data centers?

Mimi

The science works. The engineering challenges are significant but not impossible. We're probably years away from seeing it deployed at scale, but the momentum is real.

Mark

And if it does scale, what changes?

Mimi

Everything about how we think about storage costs and environmental impact. Data centers consume enormous amounts of electricity. Even replacing a portion of their storage with DNA-based systems would reshape the industry.

  • Data centers already consume 1 to 2 percent of global electricity, and that appetite grows each year — making the energy cost of storing information a genuine civilizational pressure.
  • DNA memory sidesteps the core inefficiency of silicon: because the molecule is chemically stable, encoded data persists without any continuous power supply, cutting energy demands by a factor of one hundred.
  • The technology also promises archival durability far beyond magnetic tape or solid-state drives, making it attractive to libraries, governments, and institutions that need centuries-scale preservation.
  • Speed and scale remain the critical barriers — reading and writing DNA data is slow, costs per unit are high, and no industrial infrastructure yet exists to deploy it beyond the laboratory.
  • Researchers have cleared the foundational hurdle: proof of concept is real, working, and reproducible — the race now is engineering, not physics.

In the space where biology meets computation, researchers have built a memory device from DNA that consumes a hundredth of the power required by conventional silicon storage — a development that arrives precisely as data centers strain global electricity grids. The same molecules that carry the instructions of life have been repurposed to carry digital information, persisting without any electrical current at all. It is an early but working demonstration that the long human project of storing knowledge may be entering a new chapter, one written not in silicon but in the language of living things.

Somewhere between a biology lab and a computer science department, researchers have built a memory device from DNA that runs on a hundredth of the power conventional storage requires. The timing is pointed: data centers have become industrial electricity consumers, their servers and cooling systems straining grids and budgets at global scale. DNA offers a different logic entirely.

Unlike silicon-based memory, which requires constant electrical current to hold data, DNA is a stable molecule. Information encoded in its structure simply persists — no power flowing, no data lost. Researchers have learned to write digital information into DNA sequences and read it back reliably, creating what amounts to a biological hard drive. A hundredfold reduction in power consumption is not incremental progress; it is a different kind of storage altogether.

The appeal extends beyond energy. DNA is resilient across time in ways that magnetic tape and solid-state drives are not. Engineered DNA storage systems could maintain data integrity far longer than conventional media, making them particularly attractive for the kind of archival preservation that institutions need but rarely find.

The obstacles are substantial. Reading DNA data is slow — nothing like the instant access users expect from everyday devices. Writing is complex. Costs remain high. The industrial infrastructure to manufacture and deploy DNA storage at scale does not yet exist. These are engineering problems without clear solutions.

What has been demonstrated, however, is that the underlying chemistry works. DNA stores data reliably. It can be written and read. The energy savings are real. Whether the technology can move from laboratory proof to practical deployment remains the open question — but it is a question grounded in working results, not speculation.

In a laboratory somewhere between biology and computer science, researchers have engineered something that sounds like science fiction but works like basic chemistry: they've built a memory device from DNA that demands a hundredth of the power conventional storage systems require. The breakthrough arrives at a moment when data centers have become industrial consumers of electricity, their cooling systems and servers drawing power at scales that strain grids and budgets alike. This new approach uses the same biological molecules that carry genetic instructions to instead carry digital information—a repurposing that cuts energy demands dramatically.

The fundamental insight is straightforward once you understand it: silicon-based memory, the technology that has dominated computing for decades, requires constant electrical current to maintain data. It's power-hungry by design. DNA, by contrast, is a stable molecule. Information encoded in its structure doesn't need electricity to persist. A strand of DNA sitting in a test tube will hold its data for years without a single watt of power flowing through it. Researchers have learned to write data into DNA sequences and read it back out again, creating a storage medium that is, in essence, a biological hard drive.

The implications ripple outward quickly. Data centers worldwide consume roughly 1 to 2 percent of global electricity, a figure that grows each year as cloud storage expands and artificial intelligence systems demand more computational power. Companies spend billions cooling their server farms. If DNA-based storage could replace even a fraction of conventional systems, the energy savings would be substantial—not just economically, but environmentally. A hundredfold reduction in power consumption is not a marginal improvement; it's a fundamental shift in how information could be stored and preserved.

Beyond the immediate energy question lies another appeal: longevity. DNA is resilient. Genetic material has survived in amber for millions of years. While engineered DNA storage systems won't last quite that long without careful preservation, they can maintain data integrity far longer than magnetic tape or solid-state drives, which degrade over decades. This makes DNA storage particularly attractive for archival purposes—the kind of long-term preservation that libraries, governments, and institutions need but rarely find in conventional media.

The obstacles are real, though. DNA storage is not yet fast. Reading data from DNA takes time; it's not the instantaneous access users expect from a laptop or phone. The process of writing data into DNA sequences is also slower and more complex than simply writing to flash memory. Scaling the technology from laboratory demonstrations to data center deployment requires solving engineering problems that don't yet have clear solutions. The cost per unit of storage remains high. And the infrastructure to manufacture, handle, and access DNA storage at industrial scale doesn't yet exist.

What researchers have demonstrated, however, is proof of concept. They've shown that the fundamental physics and chemistry work. DNA can store data reliably. It can be read and written. The energy demands are genuinely lower. The question now is whether the next phase of development—moving from the lab to practical systems—can overcome the speed and scalability challenges. If it does, the way we think about data storage could shift as profoundly as it did when solid-state drives began replacing mechanical hard drives. For now, the technology remains in the realm of promise, but it's a promise grounded in working demonstrations rather than speculation.

Quer a matéria completa? Leia o original em Google News ↗
Fale Conosco FAQ