This is the authors' abstract. We don't add key points for this paper.
The need for high-speed, energy-efficient computing in machine learning and communications necessitates innovations beyond conventional electronics to sustain computational power advances without requiring prohibitive energy amounts. Photonics have emerged in various applications, demonstrating significant highlights in optical linear transformations, while if successfully employed can also be used in nonlinear processes and matching functionalities. Towards this we demonstrate nonlinear optical vector processing in the form of hamming distance calculations and content addressable memory banks, using linear silicon photonic circuits at the high-speed of 50 Gb/s, enabling pattern matching and look-up operations. The processor employs a 4×4 crossbar architecture utilizing silicon germanium modulators computing hamming distance between 2-bit optical vectors. It achieves error-rates of ~10⁻³ in binary/ternary content matching, improving state-of-the-art demonstration speeds by >2.5×. Scalability is enhanced through space-wavelength multiplexing via a wavelength-division multiplexing cell, experimentally demonstrated at 50 Gb/s, offering increased computational capacity with reduced insertion losses and power consumption.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Moschos et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: