This presentation addresses the following topic(s):
- When is cloud more efficient than on-premises computing, and vice versa? How should these hybrid environments be managed? (Alternately: Discuss the efficiencies and inefficiencies of operating all on-prem, all cloud, and hybrid.)
- Machine learning and HPC: Each worthy of its own investment, or better together? (Alternately: Marriage for the ages, or heading for divorce?)
You've probably heard that public cloud can enable large scale simulation, processing, or machine learning at supercomputer scales, for anyone. You may have even experienced this yourself! A pilot or proof-of-concept (POC) might get you started. But what about the longer term - how do you operate this? What are the differences in operating a HPC cluster on-premise, or a HPC cluster in the public cloud? AND, how is this further complicated when the architecture is hybrid (between on-premise and cloud), and even includes non-traditional HPC workloads like Machine Learning or AI? Inspired by real customer stories and experiences from running HPC & ML/AI projects and operations at scale, in production, over multiple years - we share some ideas, strategies, tips, and follow-on content for operating HPC in the public cloud (and in particular, with Azure HPC).
