Photonic integrated circuit (silicon photonics)
Realizes: matrix-vector multiplication / unitary linear transforms (for neural network inference)
Arrays of Mach-Zehnder interferometers (MZIs) and microring resonators on a silicon chip implement programmable unitary matrices in the optical domain. Light encodes values as amplitude or phase; passing through a mesh of beam-splitters (MZIs) with tunable phase shifters multiplies an optical input vector by the weight matrix in a single forward pass. Because photons travel at c and interference is intrinsically parallel, a single matrix-vector multiply completes in picoseconds with energy consumption set only by modulation and detection, not arithmetic logic. MIT demonstrated a photonic processor running all key deep-learning operations on-chip. Neuromorphic silicon photonics has achieved 50 GHz tiled matrix multiplication. Speed: picoseconds per matrix-vector multiply; 50 GHz demonstrated. Capacity: 64×64 to 512×512 unitary matrices on current chips; ~4-6 bit precision.
Examples
MIT — Photonic processor for ultrafast AI computations (2024)
Fully integrated photonic processor performing all key deep neural network operations on-chip at the speed of light
Photonic matrix multiplication lights up photonic accelerator (Light: Sci. & Apps., 2022)
Review of MZI-mesh, microring-resonator, and wavelength-division-multiplexing methods for optical matrix multiplication