When you think about Kubernetes, clusters of CPU and memory resources all scaling to meet the demands of container workloads probably springs to mind. But where does GPU acceleration fit in this ...