Abstract | ||
---|---|---|
To address recent many-core architecture design, HPC applications are exploring hybrid parallel programming, mixing MPI and OpenMP. Among them, very few large scale applications in production today are exploiting asynchronous parallel tasks and asynchronous multithreaded communications to take full advantage of the available concurrency, in particular from dynamic load balancing, network, and memory operations overlapping. In this paper, we present our first results of ML-FMM algorithm implementation using GASPI asynchronous one-sided communications to improve code scalability and performance. On 32 nodes, we show an 83.5% reduction on communication costs over the optimized MPI+OpenMP version. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1007/978-3-030-22741-8_47 | COMPUTATIONAL SCIENCE - ICCS 2019, PT II |
Keywords | Field | DocType |
CEM, MLFMM, MPI, PGAS, Tasks | Architecture design,Asynchronous communication,Computer science,Concurrency,Parallel computing,Fast multipole method,Partitioned global address space,Dynamic load balancing,Distributed computing,Scalability | Conference |
Volume | ISSN | Citations |
11537 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Nathalie Möller | 1 | 0 | 0.34 |
Eric Petit | 2 | 58 | 12.73 |
Quentin Carayol | 3 | 0 | 0.34 |
Quang Dinh | 4 | 46 | 5.69 |
William Jalby | 5 | 733 | 148.24 |