Benchmarks
These benchmarks were created using the software vbench by Wes McKinney. They were run on a Intel(R) Xeon(R) CPU 5140 @ 2.33GHz on 32-Bit Linux with Python 2.6, numpy 1.6.1, and scipy 0.10.1. Benchmarks are for the pure python version, i.e. not with compiled external C extensions.
To recreate the benchmarks on your own computer, run dev/speedtracking/run_benchmarks.py -- see the comments in that file. You find the code run for the benchmarks in the same folder, additional benchmarks can be easily added following the example of the other benchmarks. The list below is automatically generated by dev/speedtracking/publish_benchmarks.py, any changes to this page after "List of benchmarks" will therefore be overridden.
List of benchmarks
Last update: 2012-05-19
- Exponential STDP (Python only)
- MultipleSpikeGeneratorGroup with list of lists
- SpikeGeneratorGroup with (index, time) pairs
- SpikeGeneratorGroup with a big array of indices/times
- SpikeGeneratorGroup with an array of indices/times
- SpikeGeneratorGroup without any spikes
- dense connection matrix (100x100)
- dense connection matrix (100x100) with weave
- dense connection matrix (100x100) with weave + codegen
- dense connection matrix (10x10)
- dynamic connection matrix (50x50)
- dynamic connection matrix (50x50) with weave
- dynamic connection matrix (5x5)
- sparse connection matrix (100x100)
- sparse connection matrix (100x100) with weave
- sparse connection matrix (100x100) with weave + codegen
- sparse connection matrix (10x10)
