5 papers
Modeling and Analysis of Application Interference on Dragonfly+
Yao Kang, Xin Wang, Neil McGlohon +3
Dragonfly class of networks are considered as promising interconnects for next-generation supercomputers. While Dragonfly+ networks offer more path diversity than the original Drag…
Union: An Automatic Workload Manager for Accelerating Network Simulation
Xin Wang, Misbah Mubarak, Yao Kang +2
With the rapid growth of the machine learning applications, the workloads of future HPC systems are anticipated to be a mix of scientific simulation, big data analytics, and machin…
Q-adaptive: A Multi-Agent Reinforcement Learning Based Routing on Dragonfly Network
Yao Kang, Xin Wang, Zhiling Lan
High-radix interconnects such as Dragonfly and its variants rely on adaptive routing to balance network traffic for optimum performance. Ideally, adaptive routing attempts to forwa…
MRSch: Multi-Resource Scheduling for HPC
Boyang Li, Yuping Fan, Matthew Dearing +4
Emerging workloads in high-performance computing (HPC) are embracing significant changes, such as having diverse resource requirements instead of being CPU-centric. This advancemen…
Study of Workload Interference with Intelligent Routing on Dragonfly
Yao Kang, Xin Wang, Zhiling Lan
Dragonfly interconnect is a crucial network technology for supercomputers. To support exascale systems, network resources are shared such that links and routers are not dedicated t…