Performance
This page gives an overview, of what to currently expect in terms of performance from Nuitka. The results are the top value from this kind of output, running pystone 1000 times and taking the minimal value. The idea is that the fastest run is most meaningful, and eliminates usage spikes. Note For the pystone benchmark, higher values are better. The numbers in parenthesis are factors, comparing to the uncompiled version. Python Uncompiled Compiled LTO Compiled PGO Debian Python 2.7 137497.87 (1.000) 460995.20 (3.353) 503681.91 (3.663) Nuitka Python 2.7 144074.78 (1.048) 479271.51 (3.486) 511247.44 (3.718)Pystone Results
echo"Uncompiled Python2"
foriin{1..100};doBENCH=1python2tests/benchmarks/pystone.py;done|sort-rn|head-n1
python2-mnuitka--lto=yes--pgotests/benchmarks/pystone.py
echo"Compiled Python2"
foriin{1..100};doBENCH=1./pystone.bin;done|sort-n|head-rn1
echo"Uncompiled Python3"
foriin{1..100};doBENCH=1python3tests/benchmarks/pystone3.py;done|sort-rn|head-n1
python3-mnuitka--lto=yes--pgotests/benchmarks/pystone3.py
echo"Compiled Python3"
foriin{1..100};doBENCH=1./pystone3.bin;done|sort-rn|head-n1