guest ::
| Home > Publications database > In Transit Coupling of Neuroscientific Simulation and Analysis on High Performance Computing Systems |
| Home > Publications database > In Transit Coupling of Neuroscientific Simulation and Analysis on High Performance Computing Systems |
| Master Thesis | FZJ-2020-00449 |
2019
Please use a persistent id in citations: http://hdl.handle.net/2128/24005
Abstract: High performance computing (HPC) is experiencing an increasing imbalance be-tween processing power and I/O capabilities. This imbalance has led to the challengeof managing the large amounts of data produced by extreme-scale simulations. Ithas become prohibitive expensive to store this data on disk for subsequent offlineanalysis. In transit processing could perform this analysis on memory-resident data.In this thesis, based on requirements of neuroscientific use cases, a frameworkhas been designed, implemented and tested on the JURECA supercomputer locatedat the Forschungszentrum Jülich. In the framework, simulation and analysis areconnected in transit across compute nodes using a client-server model. Data istransferred in a streaming manner, without disk I/O in between. The frameworkfulfills the presented use case requirements. Dedicated experiments on algorithmicsolutions of the data transfer show, that no data is lost during transfer.The design of the framework enables future integration of other software and thuscould serve as a basis for in transit coupling in neuroscientific workflows on HPCsystems.
|
The record appears in these collections: |
PDF
PDF (PDFA)