Summary:
UCLA researchers in the Department of Electrical and Computer Engineering have developed a wavelength-multiplexed, massively parallel diffractive optical system for high-density information storage, image projection, and programmable beam steering with high-fidelity reconstruction and minimal crosstalk.
Background:
The growing demands of data-intensive applications, including AI and large-scale analytics, require storage systems with high density, faster access, and longer lifetimes. Conventional magnetic storage suffers from limited speed, durability, and scalability. Optical storage, such as holographic systems, offers high-density parallel access, but practical adoption is hindered by fabrication limits, bandwidth constraints, and crosstalk that reduces reconstruction fidelity. For long-term data storage, there is a significant need for efficient and durable storage methods capable of supporting modern high-throughput computing. As data volumes continue to grow exponentially, conventional approaches struggle to meet the performance and efficiency requirements of next-generation computing and sensing systems. Thus, there remains an unmet need for an optical storage platform that combines high capacity, accurate data reconstruction, and fast, parallel readout.
Innovation:
Professor Aydogan Ozcan and his research team have developed a wavelength-multiplexed, massively parallel diffractive optical storage system, utilizing dielectric diffractive layers engineered at subwavelength scales via deep learning. Each wavelength encodes distinct image or data patterns, enabling high-density, parallel readout with minimal crosstalk. Experiments demonstrate that the system restores desired images and patterns with high fidelity across multiple spectral channels. The system thus supports high-fidelity image reconstruction, programmable beam steering with subwavelength precision, and operates across broad spectral ranges without redesign or material-specific dispersion engineering. The diffractive architecture also allows direct integration with spectral filter arrays for snapshot measurements, further extending its functionality for imaging and sensing applications. Ultimately, the system presents a versatile, scalable platform for applications including high-capacity optical information storage, image projection, and beam steering with significant benefits to storage and reconstruction fidelity over current systems.
Potential Applications:
● High-density optical information storage
● Programmable beam steering with subwavelength precision
● Holographic and volumetric displays
● Secure and private optical communication
● Snapshot spectral imaging and sensing
● Next-generation data centers
Advantages:
● High storage density
● Minimized crosstalk
● Broad spectral operation
● High-fidelity reconstruction
● Scalable and energy-efficient
State of Development:
Experimental validation of the concept completed via fabrication of two-layer diffractive device, demonstrating the storage and retrieval of six different image patterns.
Related Publications:
Che-Yung Shen, Yuhang Li, Çağatay Işıl, et al. "Wavelength-multiplexed diffractive optical information storage and image projection," Advanced Photonics 8(4), 046011 (30 Jun 2026) https://doi.org/10.1117/1.AP.8.4.046011
Press Release:
UCLA Samueli Electrical & Computer Engineering: Wavelength-multiplexed diffractive optical storage enables massively parallel image retrieval
Reference:
UCLA Case No. 2026-058
Lead Inventor:
Aydogan Ozcan, Chancellor’s Professor, Department of Electrical and Computer Engineering