Advertisement

Performance Analysis and Parallelization Strategies in Neuron Simulation Codes

  • Victor Lopez
  • Marçal Sola
  • Jesus Labarta
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8603)

Abstract

An important activity within the Human Brain Project (HBP) is to analyse and optimize the two main neuron simulation codes to improve their scalability and adapt them to efficiently work under an interactive supercomputing usage pattern. One application was already MPI + OpenMP while the other was pure MPI. We describe the analyses performed with the BSC tools and the initial efforts to hybridize the codes for better matching multicore architectures and to introduce the malleability that will be needed to operate under dynamic resource allocation environments of the future.

Keywords

Load Balance Hybrid Version Human Brain Project Hardware Counter Performance Analysis Tool 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

References

  1. 1.
    NEST Initiative, June 2014. http://www.nest-initiative.org/
  2. 2.
    NEURON|for empirically-based simulations of neurons and networks of neurons, June 2014. http://www.neuron.yale.edu/neuron/
  3. 3.
  4. 4.
  5. 5.
  6. 6.
    The OmpSs Programming Model|Programming Models@BSC, June 2014. http://pm.bsc.es/ompss/
  7. 7.
    Dynamic Load Balance|Programming Models@BSC, June 2014. http://pm.bsc.es/dlb

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  1. 1.Barcelona Supercomputing Center (BSC-CNS)BarcelonaSpain

Personalised recommendations