I have 1000s of custom (compiled to \'.so\') modules that I\'d like to use in python
at the same time. Each such module is of size (100 [KB]
) on averag
The amount of memory consumed by a single imported module is going to be at least as big as the size of the module on disk. The overhead is determined by both the OS itself (for loading a dynamic module) and Python's overhead in importing a module.
So if your module are on average 100kB in size, then importing 10000 of them will take up at least 1 GB of address space. Importing 50000 of them will run over 5 GB. You'd better be using an operating system with a 64-bit address space.