Modeling and benchmarking the potential benefit of early-bird transmission in fine-grained communication

W Schonbein, S Levy, MGF Dosanjh… - Proceedings of the …, 2023 - dl.acm.org
Traditional point-to-point communication sends data only after the entirety of the data is
available. This includes situations where multiple actors (eg, threads) contribute to the send …

A dynamic network-native MPI partitioned aggregation over Infiniband verbs

YH Temuçin, S Levy, W Schonbein… - 2023 IEEE …, 2023 - ieeexplore.ieee.org
Modern HPC systems require efficient hybrid programming model to utilize their hardware
resources effectively. The Message Passing Interface (MPI) has accommodated next …

Taking the MPI standard and the open MPI library to exascale

DE Bernholdt, G Bosilca, A Bouteiller… - … Journal of High …, 2024 - journals.sagepub.com
The Open MPI for Exascale (OMPI-X) project was one of two in the Exascale Computing
Project (ECP) focused on advancing the MPI ecosystem. The OMPI-X team worked with …

Cmb: a configurable messaging benchmark to explore fine-grained communication

WP Marts, DA Kruse, MGF Dosanjh… - 2024 IEEE 24th …, 2024 - ieeexplore.ieee.org
Modern communication APIs provide increased ability to specify when, where, and how to
send data between processes. One recent innovation is fine-grained communication, where …

Measuring Thread Timing to Assess the Feasibility of Early‐Bird Message Delivery Across Systems and Scales

WP Marts, MGF Dosanjh, W Schonbein… - Concurrency and …, 2025 - Wiley Online Library
Early‐bird communication is a communication/computation overlap technique that leverages
fine‐grained communication to improve application run‐time. Communication is divided …

[PDF][PDF] High-Performance Network-and GPU-Aware Communication for MPI Partitioned and MPI Neighbourhoods

YH Temuçin - 2024 - qspace.library.queensu.ca
Abstract Advances in High-Performance Computing (HPC) continue to improve the
performance of applications in Molecular Dynamics, AI, Deep Learning, and Large …

[PDF][PDF] Utilizing Network Hardware Parallelism for MPI Partitioned Collective Communication

YH Temuçin, AB Sedeh, W Schonbein… - Submited to 2025 32nd …, 2025 - queensu.ca
Parallel distributed applications running on largescale high-performance computing systems
depend on effective point-to-point and collective communication to meet performance goals …

[PDF][PDF] A Dynamic Network-Native MPI Partitioned Aggregation Over InfiniBand Verbs

W Schonbein, SLN Levy, R Grant, Y Temucin - 2023 - osti.gov
Modern HPC systems require efficient hybrid programming model to utilize their hardware
resources effectively. The Message Passing Interface (MPI) has accommodated next …