HPC & AI/ML in Cloud

This talk will look at the convergence of Cloud and HPC and illustrate how this intersects with AI/ML workloads. While large-scale AI/ML training jobs share many similarities with large scientific and technical HPC workloads, inference in production can benefit from elasticity and geographic distribution of Cloud. In addition, we see AI/ML entering into HPC workflows and even workloads, requiring compute environments that can cater to double floating point operations and efficient AI/ML computation in a tightly-coupled manner.