In the final project for Non-von Neumann Processors, a class that involves discussing research papers on modern computer architecture (dataflow, TPUs, systolic processors, etc.), I chose to survey a few papers and write a review on warp scheduler research. Warp schedulers are a dedicated hardware unit inside of a GPU that manages and dispatches warps (groups of 32 threads) for execution onto the GPU. They are a crucial component for GPU performance, as they are intended to hide memory latency and maximize resource utilization when scheduling groups of threads to run.