Design and optimisation of an efficient HDF5 I/O Kernel for massive parallel fluid flow simulations
| Autor: | Christoph Ertl, Jérôme Frisch, Ralf-Peter Mundani |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Gigabyte
Computer Networks and Communications Computer science Computation 010103 numerical & computational mathematics computer.file_format Hierarchical Data Format Supercomputer 01 natural sciences Computer Science Applications Theoretical Computer Science Domain (software engineering) Computational science 010101 applied mathematics File locking Computational Theory and Mathematics Kernel (image processing) 0101 mathematics Interactive visualization computer Software |
| Zdroj: | Concurrency and Computation: Practice and Experience. 29:e4165 |
| ISSN: | 1532-0626 |
| DOI: | 10.1002/cpe.4165 |
| Popis: | More and more massive parallel codes running on several hundreds of thousands of cores enter the computational science and engineering domain, allowing high-fidelity computations on up to trillions of unknowns for very detailed analyses of the underlying problems. During such runs, typically gigabytes of data are being produced, hindering both efficient storage and (interactive) data exploration. Here, advanced approaches based on inherently distributed data formats such as HDF5 become necessary in order to avoid long latencies when storing the data and to support fast (random) access when retrieving the data for visual processing. Avoiding file locking and using collective buffering, write bandwidths to a single file close to the theoretical peak on a modern supercomputing cluster were achieved. The structure of the output file supports a very fast interactive visualisation and introduces additional steering functionality. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Plný text