Synthesis of planar photonic interconnects

在过去的几年里，我们目睹了通信和计算的快速发展。由于其大带宽容量，光子学已成为几乎所有长途和城域通信网络的首选技术。相反，光子元件还无法取代计算机和局域网中的铜线。这导致计算机CPU的指令执行速度大大超过处理器到处理器和处理器到存储器链接的速度，并产生了性能瓶颈。需要一种全新的方法来克服计算和局域网中的瓶颈。互连设计和制造的最新进展清楚地表明，平面光子互连有可能改变加拿大和世界各地的信息传输和计算基础设施。该提案的中心目标是研究创建新算法以合成，模拟和可视化平面光子互连所需的基础科学。一个特别的重点将放在建立自动合成网络拓扑结构的路径，旨在实现能力，建立类似于那些已经开发的电子芯片的大规模光学电路。这将通过开发定制代码并补充有限元法求解器和光束传播算法来实现。这些工具将部署在高性能计算平台上，以综合和模拟光互连的性能。预计该项目将在加拿大培训高素质的人才。该项目的最终目标是培养新一代的专业人才，负责设计和部署新型光子互连架构。该提案预计将引起行业领导者对为接入电信市场构建光子处理器的商业兴趣。

Over the past several years, we have witnessed rapid advances in communications and computing. Photonics has emerged as the technology of choice for almost all long-haul and metro area communication networks due to its large bandwidth capacity. On the contrary, photonics components have not been able to replace copper wires in computers and local area networks. This resulted in computer CPUs having instruction execution speeds that greatly exceed the speed of processor-to-processor and processor-to-memory links by ever increasing margins, creating a performance bottleneck. A radically new approach is required to overcome bottlenecks in computing and local area networks. Recent progress in the design and manufacture of interconnects clearly shows that planar photonic interconnects have the potential to change the information transfer and computing infrastructure in Canada and worldwide. The central objective of this proposal is to research fundamental science required to create novel algorithms to synthesize, simulate, and visualize planar photonic interconnects. A special emphasis will be put on establishing the path towards automatic synthesis of net topology directed at achieving capabilities to build large-scale optical circuits similar to the ones that have been developed to build electronics chips. This will be achieved through developing custom code and supplementing it with finite element method solvers and beam propagation algorithms. These tools will be deployed on high-performance computational platforms to synthesize and simulate the performance of optical interconnects. The project is expected to train highly qualified personnel in Canada. The ultimate goal of the project is to produce a new generation of professionals that will take charge in the design and deployment of novel photonic interconnect architectures. This proposal is expected to generate commercial interest from industry leaders in building photonic processors for the access telecommunication market.