Linked Presentation: Fast, Approximate Vector Queries on Very Large Unstructured DatasetsTransparent GPU Sharing in Container Clouds for Deep Learning Workloads
