All benchmarks are platform-independent (run on any computing device with appropriate hardware). A few tests require an NVIDIA GPU with Cuda toolkit installed.
Continuous integration conducted on:
- Ubuntu (Travis-CI)
- MacOS (Travis-CI)
- Windows (AppVeyor)
Tested on Python 3.5, 3.6 and PyPy3.
This command compiles the Fortran code and prepares Python prereqs:
pip install -e .
Iterative benchmarks, here using the pisum algorithm:
Matrix Multiplication benchmarks:
hypot() is faster from 1 to a few hundred elements, then
sqrt(x^2+y^2) becomes slightly faster. However,
hypot() does not
overflow for arguments near REALMAX. For example, in Python:
from math import sqrt, hypot a=1e154; hypot(a,a); sqrt(a**2+a**2); 1.414213562373095e+154 inf
Execute the Hypot speed test by:
Julia binaries are often downloaded to a particular directory.
Python doesn't pickup
.bash_aliases, which is commonly used to point to Julia.
FC=ifort cmake ..
GNU Fortran (gfortran ≥ 6 required):
FC=gfortran cmake ..
We give a hint to CMake where your MKL libraries on. For example:
MKLROOT=/opt/intel/mkl cmake ..
Of course this option can be combined with
You can set this environment variable permanently for your convenience
(normally you always want to use MKL) by adding to your