Affiliation:
1. Barcelona Supercomputing Center, Plaça Eusebi Güell, 1-3, 08034 Barcelona, Spain
Abstract
Top-Down models are defined by hardware architects to provide information on the utilization of different hardware components. The target is to isolate the users from the complexity of the hardware architecture while giving them insight into how efficiently the code uses the resources. In this paper, we explore the applicability of four Top-Down models defined for different hardware architectures powering state-of-the-art HPC clusters (Intel Skylake, Fujitsu A64FX, IBM Power9, and Huawei Kunpeng 920) and propose a model for AMD Zen 2. We study a parallel CFD code used for scientific production to compare these five Top-Down models. We evaluate the level of insight achieved, the clarity of the information, the ease of use, and the conclusions each allows us to reach. Our study indicates that the Top-Down model makes it very difficult for a performance analyst to spot inefficiencies in complex scientific codes without delving deep into micro-architecture details.
Reference17 articles.
1. (2022, November 01). Top500 List. Available online: https://www.top500.org/lists/top500/2022/11/.
2. Roofline: An Insightful Visual Performance Model for Multicore Architectures;Williams;Commun. ACM,2009
3. Ofenbeck, G., Steinmann, R., Caparros, V., Spampinato, D.G., and Püschel, M. (2014, January 23–25). Applying the roofline model. Proceedings of the 2014 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), Monterey, CA, USA.
4. Cache-aware Roofline model: Upgrading the loft;Ilic;IEEE Comput. Archit. Lett.,2014
5. Banchelli, F., Garcia-Gasulla, M., Houzeaux, G., and Mantovani, F. (July, January 29). Benchmarking of State-of-the-Art HPC Clusters with a Production CFD Code. Proceedings of the Platform for Advanced Scientific Computing Conference, Geneva, Switzerland.