GPU-accelerated atmospheric chemical kinetics in the ECHAM/MESSy (EMAC) Earth system model (version 2.52)

Abstract

Abstract. This paper presents an application of GPU accelerators in Earth system modeling. We focus on atmospheric chemical kinetics, one of the most computationally intensive tasks in climate-chemistry model simulations. We developed a software package that automatically generates CUDA kernels to numerically integrate atmospheric chemical kinetics in the global climate model ECHAM/MESSy Atmospheric Chemistry (EMAC), used to study climate change and air quality scenarios. A source-to-source compiler outputs a CUDA-compatible kernel by parsing the FORTRAN code generated by the Kinetic PreProcessor (KPP) general analysis tool. All Rosenbrock methods that are available in the KPP numerical library are supported.Performance evaluation, using Fermi and Pascal CUDAenabled GPU accelerators, shows achieved speed-ups of 4.5× and 20.4×, respectively, of the kernel execution time. A node-to-node real-world production performance comparison shows a 1.75× speed-up over the non-accelerated application using the KPP three-stage Rosenbrock solver. We provide a detailed description of the code optimizations used to improve the performance including memory optimizations, control code simplification, and reduction of idle time. The accuracy and correctness of the accelerated implementation are evaluated by comparing to the CPU-only code of the application. The median relative difference is found to be less than 0.000000001 % when comparing the output of the accelerated kernel the CPU-only code.The approach followed, including the computational workload division, and the developed GPU solver code can potentially be used as the basis for hardware acceleration of numerous geoscientific models that rely on KPP for atmospheric chemical kinetics applications.

Enabling high-performance cloud computing for Earth science modeling on over a thousand cores: application to the GEOS-Chem atmospheric chemistry model

“…A similar pattern was shown by the Fig. 1 of Alvanos and Christoudias (2017) in their global atmospheric chemistry model. The code and further discussions for this problem are at https://github.com/geoschem/gchp/issues/44 and https://github.com/geoschem/geos-chem/issues/77.…”

Computational Benefit of GPU Optimization for the Atmospheric Chemistry Modeling

“…Therefore, the analysis here reveals that the chemistry box model performance is not limited by the computational rate but by the memory bandwidth. Since the GPU has both more floating point cores and higher memory bandwidth than the CPU (Alvanos & Christoudias, 2017;Mantell et al, 2016), we believe that it is still promising to gain some computational benefits from the GPU. Note that L2 cache misses are used as an indication of cache performance and memory bandwidth.…”

“…However, it was still possible to utilize the GPU efficiently when a large number of small matrices were solved simultaneously and independently (Abdelfattah et al, 2017;Dong et al, 2014;Haidar et al, 2018). Alvanos and Christoudias (2017) recently used the GPU accelerators to speed up the chemistry module of the global chemistryclimate model ECHAM/MESSy Atmospheric Chemistry (EMAC) by a factor of 1.75×. However, the CUDA codes in their study were parsed from the Fortran codes generated by the Kinetic PreProcessor (KPP).…”

Climate model-informed deep learning of global soil moisture distribution

“…Progress has been made to accelerate weather, climate and atmospheric-chemistry models with general purpose graphics processing units (GPUs) (Yashiro et al, 2016;Alvanos and Christoudias, 2017;Sun et al, 2018;Fuhrer et al, 2018;Müller et al, 2019), but a wider utilisation is pending. Other high performance computing applications such as lattice quantum chromodynamics with typically smaller codebase and less legacy code swiftly exploited the computational resources of GPUs (Egri et al, 2007;Clark et al, 2010).…”