Journal cover Journal topic
Geoscientific Model Development An interactive open-access journal of the European Geosciences Union
Geosci. Model Dev., 10, 2891-2904, 2017
https://doi.org/10.5194/gmd-10-2891-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.
Development and technical paper
01 Aug 2017
GNAQPMS v1.1: accelerating the Global Nested Air Quality Prediction Modeling System (GNAQPMS) on Intel Xeon Phi processors
Hui Wang1,2,4,*, Huansheng Chen2,*, Qizhong Wu1,4, Junmin Lin3, Xueshun Chen2, Xinwei Xie3, Rongrong Wang1,4, Xiao Tang2, and Zifa Wang2 1College of Global Change and Earth System Science, Beijing Normal University, Beijing 100875, China
2LAPC, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100875, China
3Intel (China) Corporation, Beijing 100013, China
4Joint Centre for Global Changes Studies, Beijing Normal University, Beijing 100875, China
*These authors contributed equally to this work.
Abstract. The Global Nested Air Quality Prediction Modeling System (GNAQPMS) is the global version of the Nested Air Quality Prediction Modeling System (NAQPMS), which is a multi-scale chemical transport model used for air quality forecast and atmospheric environmental research. In this study, we present the porting and optimisation of GNAQPMS on a second-generation Intel Xeon Phi processor, codenamed Knights Landing (KNL). Compared with the first-generation Xeon Phi coprocessor (codenamed Knights Corner, KNC), KNL has many new hardware features such as a bootable processor, high-performance in-package memory and ISA compatibility with Intel Xeon processors. In particular, we describe the five optimisations we applied to the key modules of GNAQPMS, including the CBM-Z gas-phase chemistry, advection, convection and wet deposition modules. These optimisations work well on both the KNL 7250 processor and the Intel Xeon E5-2697 V4 processor. They include (1) updating the pure Message Passing Interface (MPI) parallel mode to the hybrid parallel mode with MPI and OpenMP in the emission, advection, convection and gas-phase chemistry modules; (2) fully employing the 512 bit wide vector processing units (VPUs) on the KNL platform; (3) reducing unnecessary memory access to improve cache efficiency; (4) reducing the thread local storage (TLS) in the CBM-Z gas-phase chemistry module to improve its OpenMP performance; and (5) changing the global communication from writing/reading interface files to MPI functions to improve the performance and the parallel scalability. These optimisations greatly improved the GNAQPMS performance. The same optimisations also work well for the Intel Xeon Broadwell processor, specifically E5-2697 v4. Compared with the baseline version of GNAQPMS, the optimised version was 3.51 × faster on KNL and 2.77 × faster on the CPU. Moreover, the optimised version ran at 26 % lower average power on KNL than on the CPU. With the combined performance and energy improvement, the KNL platform was 37.5 % more efficient on power consumption compared with the CPU platform. The optimisations also enabled much further parallel scalability on both the CPU cluster and the KNL cluster scaled to 40 CPU nodes and 30 KNL nodes, with a parallel efficiency of 70.4 and 42.2 %, respectively.

Citation: Wang, H., Chen, H., Wu, Q., Lin, J., Chen, X., Xie, X., Wang, R., Tang, X., and Wang, Z.: GNAQPMS v1.1: accelerating the Global Nested Air Quality Prediction Modeling System (GNAQPMS) on Intel Xeon Phi processors, Geosci. Model Dev., 10, 2891-2904, https://doi.org/10.5194/gmd-10-2891-2017, 2017.
Publications Copernicus
Download
Short summary
We introduced some methods to port our Global Nested Air Quality Prediction Modeling System (GNAQPMS) model on Intel Knight Landing (KNL). In this paper, we introduced both common and specific methods to accelerate out model better. With the guidance of the resources material on Intel Websites (http://www.intel.com/content/www/us/en/products/processors/xeon-phi.html) and relative books, this paper could be an example for the model developers to take advantage of KNL for their model.
We introduced some methods to port our Global Nested Air Quality Prediction Modeling System...
Share