Nvidia's graphics processing units (GPUs) are the dominant chips for AI training and inference (or application deployment), ...