Abstract
The characteristics of the 5th Supercomputer Nurion Knights Landing (KNL) system of the Korea Institute of Science and Technology Information (KISTI) were analyzed by developing ultra-high resolution atmospheric and ocean numerical circulation models. These models include the Weather Research and Forecasting System (WRF), Regional Ocean Modeling System (ROMS), and Unstructured Grid Finite Volume Community Ocean Model (FVCOM). Ideal and real-case experiments were simulated for each model according to the number of parallelized cores used for comparing performances. Identical experiments were performed on a general multicore system (Skylake and a general cluster system) for a performance comparison with the Nurion KNL system. Although the KNL system has more than twice as many cores per node as the Skylake system, the KNL system demonstrated 1/3 of the performance rate of the Skylake system. However, the performance rate of the Nurion KNL system was approximately 43% for all experiments. Reducing the number of cores per node in the KNL system by half (36 cores) is the most efficient method when the total number of cores is less than 256 cores, while it is more economical to use all cores when using more than 256 cores. In all experiments, the performance was continuously improved even for a maximum core experiment (1024 cores), thereby indicating that the KNL system can effectively simulate ultra-high resolution numerical circulation models.
Highlights
The Korea Institute of Science and Technology Information (KISTI) recently installed the 5th supercomputer Nurion system, which exhibits theoretical and actual performances of 25.7 PF and13.92 PF, respectively
Cho et al (2017) found that the performance of parallel MPI with the same nodes was improved by up to 272% when using on-package memory (Multi-Channel DRAM, MCDRAM), which is one of the features of the Knights Landing (KNL) system [3]
The performance rate is defined as the sum of the simulation time of SKL and CLS divided by twice the simulation time of KNL
Summary
The white paper [2] from DELL EMS in June 2018 compared the performances of the SKL and KNL systems with several numerical models. 2.5 km resolution, and it was applied to the inputs and outputs by the parallel-netcdf 1.8.1 library for parallel optimization This configuration can improve the performance of a KNL system with many cores and low memory in distributing reading and writing model results. Yoon and Song (2019) analyzed the features of manycore architectures such as the KNL system [6] These studies were not conducted using general numerical models used in real circumstances and did not include the effects of inputs and outputs. The performance of the Nurion KNL system was compared with general cluster systems using three numerical models: the Weather Research and Forecasting System (WRF), Regional. All experiments were performed on a general cluster system (a large-scale cluster system at Jeju University, referred to as CLS) to objectively compare its performance with that of the Nurion system
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
