WDM-Enabled Multi-core Parallel Programmable Photonic Signal Processor

As a frontier in integrated photonics, programmable photonic processors based on two-dimensional waveguide meshes offer compact, reconfigurable platforms capable of performing a wide range of optical functions from spectral filtering to matrix computation, without requiring custom fabrication cycles. However, state-of-the-art Mach–Zehnder interferometer (MZI)-based architectures suffer from inherent limitations: the physical footprint of the MZI unit limits both the achievable free spectral range (FSR) and the scalability toward wavelength-division multiplexing (WDM), while parallelism relies solely on spatial routing, failing to fully exploit the wavelength domain and thus limiting functional versatility. Here, we propose and experimentally demonstrate a WDM-enabled multi-core parallel programmable photonic signal processor implemented on a reconfigurable hexagonal mesh topology, in which each MZI incorporates over-coupled microring resonators (MRRs) in both arms. This MRR-assisted MZI architecture provides wavelength-selective phase and amplitude control at the unit level, enabling a substantially higher degree of WDM parallelism while extending the programmable FSR to 1 THz and reducing the tuning power to only 4.3 mW/π. By jointly exploiting spatial and wavelength dimensions, the processor supports genuine multi-threaded execution of multiple tasks on a single chip, with seamless WDM compatibility that lies beyond the capabilities of conventional MZI-based meshes. We demonstrate the versatility of this architecture through a broad range of experiments, including flex-grid elastic WDM, tunable optical filtering, parallel optical computation, WDM-enabled image convolution, and wavelength-dependent tunable delay lines for dual-beam microwave-photonic beamforming. These results establish WDM-enabled multi-core programmable photonic processing as a promising route toward large-scale, multifunctional, and highly parallel integrated photonic systems.