Numba is a Python package that uses the LLVM compiler to compile Python code to native code. [Numba 0.13], released a few weeks ago, offers support for automatically creating CUDA kernels from Python code.
I created a [notebook][^1] ([HTML]) to show off the demo code.
[Numba 0.13]: http://continuum.io/blog/numba-0.13 “Numba 0.13”
[HTML]: http://michaelgalloy.com/wp-content/uploads/2014/04/Numba-with-CUDA.html “Numba with CUDA IPython HTML”
[notebook]: http://michaelgalloy.com/wp-content/uploads/2014/04/Numba-with-CUDA.ipynb “HTML output of Numba with CUDA IPython notebook”
[^1]: I’m not sure which is cooler, IPython notebooks or Numba.