This workshop teaches programming techniques for a range of gpu oriented tasks. After an introduction to cuda programming, participant will advance to building applications for image processing, signal processing and game programming. The training takecares of the rational of distributing an application over both CPU and GPU by implementing the techniques learned on cuda specific platform. At the end of the workshop participant will have access to all course material to exercise and practice on their own. During the training, each participant will be required to have a fully configured, GPU-equipped workstation or laptop of his own. The second option is to use Google Colaboratory for execution of cuda programs in the cloud. Upon successful completion of workshop and assessment, participants get high level understanding about vectorization, parallel computation, and heterogenous computation with cuda programming.