Efficient Direct-Connect Topologies for Collective Communications
This addresses network efficiency challenges in distributed computing systems, though it appears incremental as it builds on existing graph topologies with algorithmic improvements.
The paper tackles the problem of designing efficient network topologies for collective communications by developing an algorithmic framework that constructs direct-connect topologies optimized for latency-bandwidth trade-offs, resulting in significant performance benefits demonstrated through testbeds and simulations.
We consider the problem of distilling efficient network topologies for collective communications. We provide an algorithmic framework for constructing direct-connect topologies optimized for the latency vs. bandwidth trade-off associated with the workload. Our approach synthesizes many different topologies and schedules for a given cluster size and degree and then identifies the appropriate topology and schedule for a given workload. Our algorithms start from small, optimal base topologies and associated communication schedules and use techniques that can be iteratively applied to derive much larger topologies and schedules. Additionally, we incorporate well-studied large-scale graph topologies into our algorithmic framework by producing efficient collective schedules for them using a novel polynomial-time algorithm. Our evaluation uses multiple testbeds and large-scale simulations to demonstrate significant performance benefits from our derived topologies and schedules.