Light-driven memory could reduce energy usage for AI

As artificial intelligence, cloud computing, and digital services continue to expand, the world faces an increasing need for faster and more energy-efficient ways to store and process information. A team led by the National Institute for Quantum Science and Technology (QST) has developed a new magnetic memory material that can be rewritten using laser light instead of electrical current. This could help reduce data center power consumption and support future high-speed information systems. This study was published online applied physics letters June 8, 2026 (DOI: 10.1063/5.0328535).

This new material makes it possible to switch magnetic information with a single ultrashort laser pulse. Because light can reverse magnetic states much faster than electric current, this approach can achieve switching speeds about 1,000 times higher than traditional electrically driven magnetic memory, while also reducing heat generation and energy loss. Researchers say this advance represents a new class of low-power magnetic memory for AI hardware, edge devices, and future optoelectronic platforms.

“Today’s digital society requires faster and more sustainable memory technologies,” said Dr. Seiji Sakai, group leader at the Quantum Materials Application Research Center at the Takasaki Institute for Advanced Quantum Science, National Institutes for Quantum and Radiological Science and Technology. “By showing that practical memory materials can be switched using light, we believe this work opens a realistic path to ultra-fast, low-power devices for future information systems.”

Magnetic memory stores information by changing the direction of magnetization within a material. Existing magnetic memory technologies typically use electrical current to write data. Although this approach is attractive in that it retains information even when powered off, it also faces major limitations in that it limits write speed and increases energy consumption due to current heat generation. These challenges are becoming more acute as AI and large-scale digital infrastructure continue to drive demand for electricity.

To address this issue, the research team focused on all-optical switching, a phenomenon in which light reverses magnetic orientation without the need for an electrical current. This effect has been previously observed in ferrimagnetic materials, but those materials were not suitable for practical memories because their magnetic read properties were too weak for stable digital operation. In contrast, alloyed CoFeB is already widely used in magnetic memories because it offers nearly perfect spin polarization and excellent readout performance, but it was not considered suitable for optical switching.

Researchers have overcome that barrier by designing a new artificial ferrimagnetic material built from antiferromagnetically exchange-coupled layers of cobalt, gadolinium, and CoFeB. By tuning the thickness of each layer with atomic-scale precision and optimizing the entire multilayer structure, we created a material whose magnetic state can be reproducibly reversed with a single femtosecond laser pulse. The research team also showed that write and rewrite operations can be repeated stably, demonstrating the basic functionality required for memory applications.

“One of the most important aspects of this work is that we achieved optical switching in a CoFeB-based system that is already highly compatible with magnetic tunnel junction technology,” Sakai said. “That compatibility makes this result much more relevant to future devices than previous demonstrations that were limited to model materials.”

A key part of the research was the use of Japan’s fourth-generation synchrotron radiation facility, NanoTerasu, where the team used X-ray magnetic circular dichroism spectroscopy to analyze spin alignment and interlayer interactions within the material. These measurements provided atomic-level insight into multilayer structures and played a key role in guiding the design of new materials. The submission states that the ferrimagnet was developed using NanoTerasu’s synchrotron analysis.

The potential impact of this research extends beyond laboratory demonstrations. Faster memory and lower power consumption could help address one of the major hidden costs of the AI ​​era: the rapidly increasing power demands of data centers and advanced computing systems. In the long term, the technology could also serve as a photoelectric conversion interface between optical interconnects and electronic circuits, ultimately contributing to integrated chips that combine photonics and electronics on the same platform, according to the filing. Practical use of such materials in optoelectronic interfaces could begin within the next decade.

About the National Institute for Quantum Science and Technology
The National Institute for Quantum Science and Technology (QST) was established in April 2016 with the aim of promoting quantum science and technology in a comprehensive and integrated manner. This new organization was established by merging the National Institute of Radiological Sciences (NIRS) and some operations previously handled by the Japan Atomic Energy Agency (JAEA).

The National Institute for Quantum Science and Technology is committed to advancing quantum science and technology, building world-leading research and development platforms, and pioneering new fields that have significant academic, social, and economic impact.

About Dr. Seiji Sakai
Dr. Seiji Sakai is a researcher at the QST Quantum Materials and Applied Research Center. His research focuses on low-dimensional materials and their device applications in magnetic and spin electronics. He has published more than 130 papers on these topics and received more than 1,700 citations. Dr. Sakai is currently working on developing integrated spintronics and nanophotonics technologies to advance optoelectronic integration for next generation information systems.

reference
Doi: http://doi.org/10.1063/5.0328535