hpc_benchmark fix and cleanup #289

Merged
merged 2 commits into from Apr 4, 2016

Conversation

Projects
None yet
4 participants
@jakobj
Contributor

jakobj commented Mar 29, 2016

this PR fixes an issue that lead to poor scaling pf the hpc_benchmark.sli example. the issue was the use of [E_from E_to] Range in randomizing membrane potentials of neurons. this would allocate memory for an array containing the gids of the entire network and loop over this list on every process(!), which quickly becomes an issue for large scale benchmarks (in terms of time and memory). it is replaced by E_net GetLocalNodes.
in addition I have removed output via = which happens on every process. all output should happen via the logger, which restricts the number of processes that generate output.
I have added a rate calculation function, which was shamelessly copied from brunel-2000.sli.
minor changes in this PR:

  • removed some whitespace at end of lines
  • removed some blank lines
  • fixed indentation
  • replaced [unit] with (unit)
  • minor addition to documentation
  • changed default scale to 2 (0.2 was pointless since this amounts to a five-times fully connected network)
@jakobj

This comment has been minimized.

Show comment
Hide comment
@jakobj

jakobj Mar 29, 2016

Contributor

oh, right, forgot: I nominate @mschmidt87 and @suku248 as reviewers.

Contributor

jakobj commented Mar 29, 2016

oh, right, forgot: I nominate @mschmidt87 and @suku248 as reviewers.

@mschmidt87

This comment has been minimized.

Show comment
Hide comment
@mschmidt87

mschmidt87 Mar 30, 2016

Nice work. 👍

Nice work. 👍

@suku248

This comment has been minimized.

Show comment
Hide comment
@suku248

suku248 Apr 4, 2016

Thanks @jakobj and 👍 for merging

suku248 commented Apr 4, 2016

Thanks @jakobj and 👍 for merging

@jougs jougs merged commit e2dd701 into nest:master Apr 4, 2016

1 check passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment