-
Exploring hyper-parameter spaces of neuroscience models on high performance computers with Learning to Learn
Authors:
Alper Yegenoglu,
Anand Subramoney,
Thorsten Hater,
Cristian Jimenez-Romero,
Wouter Klijn,
Aaron Perez Martin,
Michiel van der Vlag,
Michael Herty,
Abigail Morrison,
Sandra Diaz-Pier
Abstract:
Neuroscience models commonly have a high number of degrees of freedom and only specific regions within the parameter space are able to produce dynamics of interest. This makes the development of tools and strategies to efficiently find these regions of high importance to advance brain research. Exploring the high dimensional parameter space using numerical simulations has been a frequently used te…
▽ More
Neuroscience models commonly have a high number of degrees of freedom and only specific regions within the parameter space are able to produce dynamics of interest. This makes the development of tools and strategies to efficiently find these regions of high importance to advance brain research. Exploring the high dimensional parameter space using numerical simulations has been a frequently used technique in the last years in many areas of computational neuroscience. High performance computing (HPC) can provide today a powerful infrastructure to speed up explorations and increase our general understanding of the model's behavior in reasonable times.
△ Less
Submitted 28 February, 2022;
originally announced February 2022.
-
Performance Comparison for Neuroscience Application Benchmarks
Authors:
Andreas Herten,
Thorsten Hater,
Wouter Klijn,
Dirk Pleiter
Abstract:
Researchers within the Human Brain Project and related projects have in the last couple of years expanded their needs for high-performance computing infrastructures. The needs arise from a diverse set of science challenges that range from large-scale simulations of brain models to processing of extreme-scale experimental data sets. The ICEI project, which is in the process of creating a distribute…
▽ More
Researchers within the Human Brain Project and related projects have in the last couple of years expanded their needs for high-performance computing infrastructures. The needs arise from a diverse set of science challenges that range from large-scale simulations of brain models to processing of extreme-scale experimental data sets. The ICEI project, which is in the process of creating a distributed infrastructure optimised for brain research, started to build-up a set of benchmarks that reflect the diversity of applications in this field. In this paper we analyse the performance of some selected benchmarks on an IBM POWER8 and Intel Skylake based systems with and without GPUs.
△ Less
Submitted 7 August, 2019;
originally announced August 2019.
-
A Lagrangian model for the evolution of turbulent magnetic and passive scalar fields
Authors:
T. Hater,
H. Homann,
R. Grauer
Abstract:
In this paper we present an extension of the \emph{Recent Fluid Deformation (RFD)} closure introduced by Chevillard and Meneveau (2006) which was developed for modeling the time evolution of Lagrangian fluctuations in incompressible Navier-Stokes turbulence. We apply the RFD closure to study the evolution of magnetic and passive scalar fluctuations. This comparison is especially interesting since…
▽ More
In this paper we present an extension of the \emph{Recent Fluid Deformation (RFD)} closure introduced by Chevillard and Meneveau (2006) which was developed for modeling the time evolution of Lagrangian fluctuations in incompressible Navier-Stokes turbulence. We apply the RFD closure to study the evolution of magnetic and passive scalar fluctuations. This comparison is especially interesting since the stretching term for the magnetic field and for the gradient of the passive scalar are similar but differ by a sign such that the effect of stretching and compression by the turbulent velocity field is reversed. Probability density functions (PDFs) of magnetic fluctuations and fluctuations of the gradient of the passive scalar obtained from the RFD closure are compared against PDFs obtained from direct numerical simulations.
△ Less
Submitted 8 December, 2010; v1 submitted 8 October, 2010;
originally announced October 2010.
