Defining Performance of Scientific Application Workloads on the AMD Milan Platform
Tsai-Wei Wu, Stephen Lien Harrell, Geoffrey Lentner, Alex Younts, Sam Weekly, Zoey Mertes, Amiya Kumar Maji, Preston Smith, Xiao Zhu · Practice and Experience in Advanced Research Computing · 2021
Understanding the capabilities of new architectures is key to informing system purchases and good long-term ROI (Return of Investment) for cluster installations. The newest AMD architecture, Milan, has become available first on Microsoft Azure and we use this early access to measure the performance of this 3rd Generation AMD EPYC processor. In this paper single node performance is gathered for seven popular scientific applications and benchmark test-suites. Quantitative comparisons are carried out between two independent platforms, Milan and its architectural predecessor Rome, for performance evaluations. Our results have shown that Milan architecture have improved performance and met our projections.