Quantifying Message Aggregation Optimisations for Energy Savings in PGAS Models
By:
Welch, Donald A; Hernandez Mendoza, Oscar R; Poole, Stephen; Poole, Wendy
Page Number:
1-8
Book Title:
Proceedings of the 2025 Annual Cray User Group (CUG) Conference
Publication Date:
June 2025
Publisher Location:
Cray User Group, District of Columbia, United States of America
Conference Name:
Computing Horizons: CUG 2025
Conference Location:
Jersey City, New Jersey, United States of America
Conference Sponsor:
HPE
Abstract
Upon breaking past the exascale barrier, HPC systems are facing their greatest challenge yet - a power wall that must be addressed through new methods in both hardware and software. While energy costs are becoming a major issue at all levels, of particular concern is that of the network, as the relative cost of moving data is increasing faster than ever. The partitioned global address space (PGAS) model is critical within certain HPC domains, but is known to suffer from the small message problem, where irregular many-to-many access patterns result in congesting the network with excessive numbers of small messages. To address this, the conveyor aggregation library was developed to defer individual messages and group them for subsequent bulk processing. In this paper, we investigate its impact on energy use related to the network, with a focus on the Slingshot 11 interconnect. We will demonstrate that this strategy is not only highly performant, but also crucial to reducing energy footprints to remain within target power envelopes.
