| Summary: | Advancements in optical coherence control have unlocked a plethora of cutting-edge applications, including long-haul communication, light detection and ranging, and optical coherence tomography. Prevailing wisdom suggests that using more coherent light sources leads to enhanced system performance and device functionalities. Our study introduces a photonic convolutional processing system that capitalizes on partially coherent light to boost computing parallelism without substantially sacrificing accuracy, potentially enabling larger-size photonic tensor cores. The reduction of the degree of coherence optimizes bandwidth utilization in the photonic convolutional processing system. This breakthrough challenges the traditional belief that coherence is essential or even advantageous in integrated photonic accelerators, thereby enabling the employment of light sources with less rigorous feedback control and thermal management requirements for high-throughput photonic computing. We demonstrate such a system in two photonic platforms for computing applications: a photonic tensor core using phase-change material photonic memories that delivers parallel convolution operations to classify gaits of ten Parkinson’s disease patients with a 92.2% accuracy (92.7% theoretically), and a silicon photonic tensor core with embedded electroabsorption modulators (EAM) to facilitate 0.108 tera operations per second (TOPS) convolutional processing for classifying MNIST handwritten digits dataset with a 92.4% accuracy (95.0% theoretically).
|
|---|