This title appears in the Scientific Report :
2015
Please use the identifier:
http://hdl.handle.net/2128/9076 in citations.
Modeling the I/O behavior of the NEST simulator using a proxy
Modeling the I/O behavior of the NEST simulator using a proxy
NEST is a simulator for spiking neural networks. It runs on ordinary desktop computers and notebooks, small clusters and supercomputers. Storing simulation data efficiently is essential for neuroscientific studies but is not trivial on supercomputers with centralized storage. To assess different I/O...
Saved in:
Personal Name(s): | Schumann, Till |
---|---|
Frings, Wolfgang / Peyser, Alexander / Schenck, Wolfram / Thust, Kay / Eppler, Jochen Martin (Corresponding author) | |
Contributing Institute: |
JARA - HPC; JARA-HPC Jülich Supercomputing Center; JSC |
Published in: |
Conference Proceedings of the YIC GACM 2015 / ed.: Stefanie Elgeti ; Jaan-Willem Simon |
Imprint: |
RWTh Aachen University
2015
|
Physical Description: |
4 p. |
Conference: | 3rd ECCOMAS Young Investigators Conference, Aachen (Germany), 2015-07-20 - 2015-07-23 |
Document Type: |
Contribution to a book Contribution to a conference proceedings |
Research Program: |
SimLab Neuroscience Supercomputing and Modelling for the Human Brain Theory, modelling and simulation Computational Science and Mathematical Methods |
Link: |
Get full text OpenAccess |
Publikationsportal JuSER |
NEST is a simulator for spiking neural networks. It runs on ordinary desktop computers and notebooks, small clusters and supercomputers. Storing simulation data efficiently is essential for neuroscientific studies but is not trivial on supercomputers with centralized storage. To assess different I/O strategies and libraries, we have implemented a proxy which imitates the writing behavior of NEST. This proxy is useful in benchmarking and statistical analysis, and thus consequent optimization, without the complexity of running full NEST simulations. |