Since the genesis of the cloud, datacenter network (DCN) infrastructure has become ubiquitous across the globe and is expected to support highly customized, fine-grained applications (e.g., HPC, distributed ML, DNN, etc.) with stringent performance requirements. However, as we move to the “post-Moore’s law era” of networking, CMOS-based electrical packet-switch ASICs are struggling to cope up with the increasing capacity while maintaining low power consumption and cost. This energy-critical situation has led to several recent proposals regarding all-optical circuit-switched DCN core design for sustainable future-generation clouds. Optical circuit-switching (OCS) technologies are the key components that make those proposals fundamentally promising, as OCS is inherently sustainable with the unique advantages: a) agnostic to data rate, b) negligible/zero power consumption, c) negligible forwarding latency, and d) no need for a frequent upgrade. However, existing round-robin based OCS cores perform poorly under realistic workloads, having high traffic skewness and high volume of inter-rack traffic. To address this issue, we propose a novel DCN architecture OSSV: a combination of OCS-based core (between ToR switches) and OCS-based reconfigurable edge (between servers and ToR switches). On one hand, the OCS core is traffic agnostic and realizes reconfigurably non-blocking ToR-level connectivity. On the other hand, OCS-based edge reconfigures itself to reshape the incoming traffic to jointly minimize traffic skewness and inter-rack traffic volume. We design a novel optimization framework: Skewness and Volume (SV) minimization, that finds the right balance between these intertwined objectives. Our extensive simulations and testbed evaluation demonstrate that OSSV can significantly reduce the performance gap between all-optical DCN cores vs. non-blocking packet-switched DCN under diverse traffic scenarios, while adding a small amount of power and cost.
Dr. Sushovan Das is a Postdoctoral Researcher at ETH Zürich in the Networked Systems Group led by Prof. Laurent Vanbever. He received his PhD in Computer Science from Rice University in 2024, advised by Prof. T. S. Eugene Ng. Prior to that, he completed M.Tech. from IIT Kharagpur (E&ECE) under Prof. Debasish Datta and a B.Tech. from Jadavpur University (ETCE). His research primarily focuses on designing novel and energy-efficient optical network architectures to optimize diverse application performance in next-generation cloud infrastructures, bridging theory and system realization. His work has appeared in IEEE INFOCOM, IEEE/ACM ToN, USENIX NSDI, ACM HotNets, IEEE/Optica JLT, and IEEE ICC.