Graphics processing units (GPUs) are compute platforms that are ideal for highly parallel workloads due to their high degree of hardware parallelism. Parallelism offered by GPUs lends itself well to machine learning and ...