Light could flow directly from communication to storage, with minimal energy loss
At the University of Augsburg, a German-Japanese research team has demonstrated for the first time that light alone can write and retrieve information in antiferromagnetic materials, bypassing the electric currents and magnetic fields that have long defined data storage. The work, published in Nature Materials, resolves a decades-old impasse: antiferromagnets were known to be faster and more resilient than conventional storage media, yet no reliable method existed to control them. By using the direction of ultrashort laser pulses rather than their polarization, the team has opened a path toward storage systems that speak the same language as global fiber-optic networks. It is a moment when two mature technological worlds — optical communication and magnetic storage — discover they may, at last, become one.
- The promise of antiferromagnetic storage has been stalled for years by a single stubborn problem: no one could reliably write information into these materials without electric currents or magnetic fields.
- A German-Japanese team has now broken that impasse, using the travel direction of ultrashort laser pulses to switch magnetic states and encode data optically for the first time.
- The method operates at standard telecommunications wavelengths, meaning it could plug directly into existing fiber-optic infrastructure without costly signal conversion.
- Written magnetic patterns persist after the laser is switched off, confirming the nonvolatility essential for any real-world storage application.
- The convergence points toward data centers that process information at the speed of light while consuming a fraction of their current electricity — a meaningful pressure valve as AI and cloud computing drive global energy demand upward.
A research team from Germany and Japan has solved one of the central puzzles blocking next-generation data storage: how to write information into antiferromagnetic materials using nothing but light. Led by physicist István Kézsmárki at the University of Augsburg, the work abandons electric currents and magnetic fields entirely, encoding data through ultrashort laser pulses alone. The results appear this week in Nature Materials.
Antiferromagnets have long been recognized as ideal candidates for faster, more resilient storage — they respond with exceptional speed and shrug off external interference. The obstacle was always control. Without a reliable way to manipulate their internal magnetic states, the theoretical advantages stayed theoretical.
Kézsmárki's team found their opening by shifting focus from light's polarization to its direction of travel. Aiming laser pulses at specific angles and intensities, they could switch the material between magnetic states — writing data in — and read it back out again through the same optical method. Crucially, the written patterns proved stable after the light was extinguished, satisfying the nonvolatility requirement that any practical storage technology demands.
The method operates at telecommunications wavelengths already embedded in global fiber-optic networks, raising the prospect of merging optical communication and magnetic storage at the material level — data flowing from network to storage at the speed of light, with no conversion losses between optical and electrical signals.
The energy stakes are considerable. Data centers already account for one to two percent of global electricity consumption, a share climbing alongside artificial intelligence and cloud computing. A storage architecture built on optical pulses rather than electrical currents could reduce that burden substantially while delivering speed gains that conventional magnetic media cannot match. The road from laboratory proof to commercial device remains long, but the foundational barrier has been cleared.
A research team spanning Germany and Japan has cracked a problem that has stalled the development of next-generation data storage: how to write information into antiferromagnetic materials using nothing but light. The breakthrough, led by experimental physicist István Kézsmárki at the University of Augsburg, dispenses entirely with electric currents and magnetic fields—the traditional tools for manipulating magnetic data. Instead, the team uses ultrashort laser pulses to encode information directly into the material's magnetic structure, a feat published this week in Nature Materials.
Antiferromagnets have long been recognized as the ideal foundation for faster, more resilient storage devices. These materials respond with exceptional speed and show remarkable resistance to external interference—qualities that make them far superior to the magnetic materials that power today's hard drives and solid-state storage. The catch has always been control. Until now, scientists lacked a reliable way to precisely manipulate the magnetic states within antiferromagnetic materials, which meant the theoretical advantages remained largely theoretical.
Kézsmárki's team solved this by abandoning the conventional approach. Rather than relying on the polarization of light—the orientation of its electromagnetic waves—they used the direction in which the laser pulse travels through the material. This seemingly subtle shift opens an entirely new pathway for writing and reading magnetic information. By directing laser light at specific angles and intensities, the researchers can switch the material between different magnetic states, effectively writing data into it. The same optical method allows them to read the information back out, creating a completely light-based storage system.
The practical implications are substantial. The team demonstrated that their method operates at telecommunications wavelengths, the same frequencies already embedded in global fiber-optic networks. This compatibility means that in principle, optical communication systems and magnetic data storage could be directly integrated—no conversion between light and electrical signals required. Data could flow from one system to another at the speed of light, with minimal energy loss in translation.
Equally important, the researchers proved that the magnetic patterns they write into the material remain stable over time. Information written with laser pulses persists even after the light is switched off, a property called nonvolatility that is essential for any practical storage technology. By repeatedly cycling the material with laser light, they confirmed that data can be reliably stored and retrieved indefinitely.
The energy implications are difficult to overstate. Data centers consume roughly 1 to 2 percent of global electricity, and that fraction is growing as artificial intelligence and cloud computing expand. Every step in the storage and retrieval process—writing data, maintaining it, reading it back—consumes power. A system that eliminates electrical signals entirely, replacing them with optical pulses, could reduce that consumption dramatically. The speed gains would be equally significant: light travels faster than electrical current through conventional circuits, and the antiferromagnetic materials themselves respond more quickly than conventional magnetic storage media.
The work represents a convergence of two previously separate technological domains. Optical communication has matured into a global infrastructure. Magnetic storage has evolved into a mature but energy-intensive technology. What Kézsmárki's team has shown is that these two worlds can be merged at the material level, creating devices that are simultaneously faster, more efficient, and more resilient to interference. The path from laboratory demonstration to commercial product remains long, but the fundamental barrier—proving that optical control of antiferromagnetic states is possible—has been cleared.
Notable Quotes
The team demonstrated that complex magnetic patterns can be selectively written into the material and stored stably through repeated laser switching.— Nature Materials publication
The Hearth Conversation Another angle on the story
Why does the direction of the laser pulse matter more than its polarization? That seems like a small distinction.
It's actually the difference between having a light switch and having a dimmer. Polarization is like the color of the light—it's a property of the wave itself. The direction of propagation is about where the light is going. When you use direction, you're essentially using the geometry of how the light moves through the material to trigger the magnetic switching. It's more precise and, it turns out, more controllable.
And antiferromagnets are better than regular magnets for storage because...?
Speed and stability. A regular magnet in a hard drive is like a compass needle—it points in one direction and external magnetic fields can nudge it around. Antiferromagnets are two magnetic systems pointing in opposite directions, canceling each other out externally. They switch states much faster and they don't care if you wave a magnet near them. That's huge for reliability.
So the real revolution here is that you're using light instead of electricity to write the data?
Exactly. Right now, to write data magnetically, you send electrical current through a coil to generate a magnetic field. That takes power and time. With this method, you're using photons—particles of light—to do the same job. Light is faster and, in principle, more efficient because you're not generating heat through electrical resistance.
The fact that it works at telecommunications wavelengths—is that just convenient, or does it fundamentally change what's possible?
It's transformative. It means you could theoretically build a data center where the same fiber-optic cables that carry your internet traffic also write and read data from storage. No conversion layers, no translation between optical and electrical domains. That's where the real energy savings come from.
How far away is this from actually being in a device you could buy?
The physics is proven. The next phase is engineering—making it reliable, scaling it up, integrating it with actual storage architectures. That's typically a five-to-ten-year journey from lab to product. But the hard part, the part that could have failed, is already done.