Photonic Chip With A 2D Programmable Waveguide (Boston U., UC Irvine, Yale)
Researchers from Boston University, UC Irvine, and Yale University have published a technical paper on "Arbitrary control over multimode wave propagation for machine learning." They introduced a device with a 2D programmable waveguide that allows for rapid reprogramming of wave propagation, enabling neural network inference without digital pre-processing or post-processing. The device offers a reduction in device area and scaling benefits for photonic processors, potentially revolutionizing photonic chip technology. The study showcases the potential of two-dimensional programmable waveguides in enhancing computational efficiency and performance in machine learning applications.
SemiEngineering•