1. Torsten Becker
  2. pep-393-bench



Naive PEP 393 Micro Benchmarks

This repository hosts a set of micro benchmarks to evaluate the PEP 393 implementation.


Create folders with different Python builds, maybe adjust measure.py with the locations of these and then run ./measure.py:

[ ... lots of output, followed by something like this: ]

=== Results ===
 access.py with pep-393               : CPU_min=2.81211 MEM=6400 KiB
 access.py with pep-393-release       : CPU_min=0.81904 MEM=5052 KiB
 access.py with py3k-default-debug    : CPU_min=2.69240 MEM=6608 KiB
 access.py with py3k-default-release  : CPU_min=0.86020 MEM=5120 KiB
 concat.py with pep-393               : CPU_min=0.42717 MEM=6328 KiB
 concat.py with pep-393-release       : CPU_min=0.19980 MEM=4976 KiB
 concat.py with py3k-default-debug    : CPU_min=0.46501 MEM=6464 KiB
 concat.py with py3k-default-release  : CPU_min=0.22005 MEM=4956 KiB
   iter.py with pep-393               : CPU_min=2.09208 MEM=6404 KiB
   iter.py with pep-393-release       : CPU_min=0.39217 MEM=5048 KiB
   iter.py with py3k-default-debug    : CPU_min=1.91860 MEM=6604 KiB
   iter.py with py3k-default-release  : CPU_min=0.38138 MEM=5120 KiB
   join.py with pep-393               : CPU_min=3.73133 MEM=8584 KiB
   join.py with pep-393-release       : CPU_min=1.89221 MEM=7180 KiB
   join.py with py3k-default-debug    : CPU_min=7.42312 MEM=10768 KiB
   join.py with py3k-default-release  : CPU_min=3.85390 MEM=9208 KiB