Researchers Store Data in Three Dimensions Using AI and Holography

A research team at Fujian Normal University in China has developed a new holographic storage technique that uses AI to store and retrieve data across three dimensions of light simultaneously: amplitude, phase, and polarization.

Traditional holographic storage uses one or two properties of light. The new method combines all three, dramatically increasing storage density within the same physical volume. A convolutional neural network model enables the use of polarization as an independent information dimension and decodes all three dimensions simultaneously from intensity images.

The research was published in the journal Optica (Optica Publishing Group) on March 26, 2026. The team, led by Xiaodi Tan, demonstrated that the method simplifies readout while significantly increasing information density.

The practical implications could be substantial: smaller data centers, more efficient large-scale archival storage, and faster data transmission. The technique could also contribute to safer data transmission, optical encryption, and advanced imaging.

At a time when global demand for data storage is growing exponentially, this represents a breakthrough that combines fundamental optics with modern deep learning in a way that could reshape how we think about physical storage.