2020
DOI: 10.1007/s11390-020-9822-9
|View full text |Cite
|
Sign up to set email alerts
|

ExaHDF5: Delivering Efficient Parallel I/O on Exascale Computing Systems

Abstract: Scientific applications at exascale generate and analyze massive amounts of data. A critical requirement of these applications is the capability to access and manage this data efficiently on exascale systems. Parallel I/O, the key technology enables moving data between compute nodes and storage, faces monumental challenges from new application, memory, and storage architectures considered in the designs of exascale systems. As the storage hierarchy is expanding to include node-local persistent memory, burst bu… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
9
0

Year Published

2020
2020
2025
2025

Publication Types

Select...
5
3
2

Relationship

4
6

Authors

Journals

citations
Cited by 38 publications
(17 citation statements)
references
References 10 publications
0
9
0
Order By: Relevance
“…For example, a recent pilot study in computational fluid dynamics showed that it could be more than 200 times faster than the same workload on an optimized number of cores on the NETL's supercomputer JOULE 2.0 [Rocki et al, 2020]. Similar scaling performance has been reported on other exascale computing clusters involving hundreds of GPU's [Byna et al, 2020].…”
Section: Model Scalibilitymentioning
confidence: 70%
“…For example, a recent pilot study in computational fluid dynamics showed that it could be more than 200 times faster than the same workload on an optimized number of cores on the NETL's supercomputer JOULE 2.0 [Rocki et al, 2020]. Similar scaling performance has been reported on other exascale computing clusters involving hundreds of GPU's [Byna et al, 2020].…”
Section: Model Scalibilitymentioning
confidence: 70%
“…Hierarchical Data Format version 5 (HDF5) [24], [25] is the most popular I/O middleware and file format used at NERSC. HDF5 provides a data model, library, and file format along with a rich programming interface for storing and managing data.…”
Section: E Hdf5 Features and Improvementsmentioning
confidence: 99%
“…HDF5 has been used heavily at supercomputing facilities for storing, reading, and querying scientific datasets [16], [17]. This is because HDF5 has specific designs and performance optimizations for popular parallel file systems such as Lustre [18], [19]. Moreover, HDF5 also provides users dynamically loaded filters [20] such as lossless and lossy compression [21], which can automatically store and query data in compressed formats.…”
Section: Introductionmentioning
confidence: 99%