With so many cores available for computation, even for cellphones, programming language Python’s global interpreter lock (GIL) has caused a bit of consternation in it’s ability to run only one thread at a time. While this may not be much of a issue as most programmers are writing sequential programming code anyway. However, for number crunching (big data, everywhere), better alternative is probably to import CUDA/OpenCL libraries to Python and write short, efficient Python code. Then again, it leaves all other cores in CPU idle. Making number crunching tasks parallel isn’t too hard, as CUDA-BLAS has shown. Python and other programming languages will need to implement parallel programming paradigm as multi-core (GPUs are already thousands of ‘cores’) processing is here.