The development of HPC applications poses specific challenges that are not easily addressed by means of standard programming tools. For this reason, people working with HPC codes are offered specialized solutions to develop, profile and debug their projects. In most cases, knowledge of such tools is limited to the ones that researchers use most frequently, while carrying out their projects. To make HPC development more effective, there are specialized courses that are designed to keep the user community up to date with programming solutions and hardware technologies. However, since in many cases these courses imply access to limited resources and have strict timing requirements, it is difficult to grant wide participation. In this colloquium, I will present a summary of the contents of the recent editions of the CINECA GPU-HACKATHON and of the Argonne Training Program on Extreme Scale Computing (ATPESC), which were held between May and August 2024. The colloquium will start from a description of the course activities and of the related material and then it will give a more specific focus on some relevant technologies to develop GPU oriented code and to perform parallel profiling, debug and result visualization, with some considerations on technical advices about efficiency and portability.