Optical neural networks (ONNs) have demonstrated record-breaking potential in high-performance neuromorphic computing due to their ultra-high execution speed and low energy consumption. However, current learning protocols fail to provide scalable and efficient solutions to photonic circuit optimization in practical applications. In this work, we propose a novel on-chip learning framework to release the full potential of ONNs for power-efficient in situ training. Instead of deploying implementation-costly back-propagation, we directly optimize the device configurations with computation budgets and power constraints. We are the first to model the ONN on-chip learning as a resource-constrained stochastic noisy zeroth-order optimization problem, and propose a novel mixed-training strategy with two-level sparsity and power-aware dynamic pruning to offer a scalable on-chip training solution in practical ONN deployment. Compared with previous methods, we are the first to optimize over 2,500 optical components on chip. We can achieve much better optimization stability, 3.7x-7.6x higher efficiency, and save >90% power under practical device variations and thermal crosstalk.
翻译:光学神经网络(ONNs)由于超高执行速度和低能消耗,在高性能神经形态计算中表现出破纪录的潜力。然而,目前的学习协议未能为实际应用中的光电路优化提供可扩展的高效解决方案。在这项工作中,我们提议了一个新型的芯片学习框架,以释放ONNs在现场高能效培训方面的全部潜力。我们不采用成本成本成本的反向调整,而是用计算预算和电力限制来直接优化设备配置。我们是第一个将ONN在芯片上学习作为受资源限制的受资源限制的随机扰动零序优化问题的模型,并提出一个新的混合培训战略,配有两级的宽度和能能能动性动态调整,以便在实际部署ONN时提供可扩展的芯片培训解决方案。与以前的方法相比,我们是第一个在芯片上优化超过2 500个光学组件的方法。我们可以实现更好的优化稳定性,3.7x-7.6x更高的效率,并在实际设备变化和热交叉跟踪下节省超过90%的功率。