UCLA Breakthrough: Optical System Achieves Fast Nonlinear Computation
URGENT UPDATE: Researchers at the University of California, Los Angeles (UCLA) have achieved a groundbreaking breakthrough in optical computing, enabling the rapid execution of large-scale nonlinear computations using only linear materials. Published in eLight, this study reveals that diffractive optical processors can simultaneously compute numerous nonlinear functions with unprecedented speed and efficiency.
This innovative system is poised to revolutionize fields reliant on nonlinear operations, including machine learning, pattern recognition, and general-purpose computing. For the first time, UCLA’s research team has demonstrated that nonlinear computation can emerge from purely linear optical interactions by encoding input variables into the phase of an optical wavefront, processed through a compact, static diffractive optical architecture.
“By exploiting diffractive processing and wavefront encoding, we have unlocked a powerful class of optical systems that compute nonlinear functions at large scale and speed with a massive spatial density,” stated Aydogan Ozcan, Chancellor’s Professor of Electrical and Computer Engineering at UCLA and the lead author of the study.
The researchers have established that these diffractive processors serve as universal nonlinear function approximators, capable of realizing any arbitrary set of band-limited nonlinear functions, including complex-valued and multivariate functions. They also successfully approximated typical nonlinear activation functions used in digital neural networks, such as sigmoid, tanh, ReLU (rectified linear unit), and softplus.
In a remarkable demonstration, numerical simulations revealed the capability to compute one million distinct nonlinear functions at wavelength-scale spatial density, executed through an optimized diffractive optical processor. The architecture was experimentally validated with a compact optical setup that included a spatial light modulator and an image sensor, successfully learning and executing tens of distinct nonlinear functions simultaneously.
This framework is scalable, with potential advancements in ultrafast analog computing, neuromorphic photonics, and high-throughput optical signal processing. By utilizing high-end image sensors with hundreds of megapixels, the system could compute hundreds of millions of nonlinear functions in parallel—achieved without the need for nonlinear optical materials or electronic post-processing.
As this technology progresses, it may significantly impact industries reliant on fast data processing and complex computations, paving the way for advancements in artificial intelligence and beyond. Stay tuned for further updates as UCLA continues to push the boundaries of optical computing technology.
