Responsibilities may be quite diverse of a nonexempt technical nature. U.S. experience and education requirements will vary significantly depending on the unique needs of the job. Job assignments are usually for the summer or for short periods during breaks from school. Developer tools and analyzers- The project is to develop tools and analysis for GPU performance in the context of High-Performance Computing (HPC) and AI. In particular the analysis is: Full characterization of Compute Kernels running on GPUs. The data of interest is huge but must be collected in partial form in single iteration due to hardware limitations and thus requires hundreds of such iterations. Software analysis via lean code instrumentations to record memory access patterns. Such data generation may severely affect performance and must be made apart from the HW collection. The target of collection and analysis is of a GPU Kernel that will be CAPTURED and REPLAYED multiple times to produce the above-mentioned data over hundreds of runs. Another variable to control in the experiment are variations on target HW.