Inter-APU Communication on AMD MI300A Systems via Infinity Fabric: A Deep Dive
Open Access
- 6 October 2025
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 238-250
Abstract
No abstract availableAll Related Versions
Funding Information
- European High Performance Computing Joint Undertaking (101093261)
- Swedish Research Council (2022.03062)
This publication has 23 references indexed in Scilit:
- Dissecting CPU-GPU Unified Physical Memory on AMD MI300A APUs2025
- Realizing the AMD Exascale Heterogeneous Processor Vision : Industry Product2024
- Multi-GPU UNRES for scalable coarse-grained simulations of very large protein systemsComputer Physics Communications, 2024
- High Performance MPI over the Slingshot InterconnectJournal of Computer Science and Technology, 2023
- Tartan: Evaluating Modern GPU Interconnect via a Multi-GPU Benchmark Suite2018
- Performance Modeling and Evaluation of Distributed Deep Learning Frameworks on GPUs2018
- NUMA-Aware Data-Transfer Measurements for Power/NVLink Multi-GPU Systems2018
- Evaluating On-Node GPU Interconnects for Deep Learning Workloads2017
- Design and Analysis of an APU for Exascale Computing2017
- On the Performance Portability of Structured Grid Codes on Many-Core Computer Architectures2014