Recent advances in the hardware development allow us to program wider range of devices - but those devices must be programmed using different programming models. There are libraries that can help us to program GPUs or FPGAs (for example OpenCL or CUDA) but they are written with C language in mind and focus on low-level details.
Fortunately there are Python wrappers that take care of all the boring details allowing us to focus on the interesting aspects of programming GPU. This talk will introduce you to programming GPUs with the help of PyOpenCL and PyCUDA, and present techniques allowing you to get the most from your graphics card. After it, you will be able to run code both on your CPU and GPU!