How to Speed Up Build Time ?¶
For projects using SCons or developers using build systems in IDEs (see the How to Configure my IDE ? section), it is possible to speed up the build time by parallelizing most of C++ compilations.
This is usually done using the
-j<CPU_COUNT> flag with SCons:
will use, when possible, \(6\) concurrent compilations.
It is not recommended to set
<CPU_COUNT> to a value superior to your number of processors.
From a console, you can see yout number of processors by typing the following command line in a Python interpreter:
import multiprocessing multiprocessing.cpu_count()
To avoid using the
-j<CPU_COUNT> flag each time, you can type the following command line in your console:
This will automatically set the number of concurrent compilations to your number of processors minus one.
You can manually specify the number of concurrent compilations using the
--number <CPU_COUNT> flag.