uMMAP-IO: User-Level Memory-Mapped I/O for HPC

The integration of local storage technologies alongside traditional parallel file systems on HPC clusters, is expected to rise the programming complexity on scientific applications aiming to take advantage of the increased-level of heterogeneity. In this work, we present uMMAP-IO, a user-level memor...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE 26th International Conference on High Performance Computing, Data, and Analytics (HiPC) pp. 363 - 372
Main Authors Rivas-Gomez, Sergio, Fanfarillo, Alessandro, Valat, Sebastien, Laferriere, Christophe, Couvee, Philippe, Narasimhamurthy, Sai, Markidis, Stefano
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2019
Subjects
Libraries
Memory management
Memory Mapped I/O
Parallel I/O
Programming
Random access memory
Resource management
Supercomputers
Synchronization
uMMAP IO
Online AccessGet full text

Cover

More Information
Summary:The integration of local storage technologies alongside traditional parallel file systems on HPC clusters, is expected to rise the programming complexity on scientific applications aiming to take advantage of the increased-level of heterogeneity. In this work, we present uMMAP-IO, a user-level memory-mapped I/O implementation that simplifies data management on multi-tier storage subsystems. Compared to the memory-mapped I/O mechanism of the OS, our approach features per-allocation configurable settings (e.g., segment size) and transparently enables access to a diverse range of memory and storage technologies, such as the burst buffer I/O accelerators. Preliminary results indicate that uMMAP-IO provides at least 5-10x better performance on representative workloads in comparison with the standard memory-mapped I/O of the OS, and approximately 20-50% degradation on average compared to using conventional memory allocations without storage support up to 8192 processes.
ISSN:2640-0316
DOI:10.1109/HiPC.2019.00051