Articles | Volume 10, issue 10
Geosci. Model Dev., 10, 3679–3693, 2017
https://doi.org/10.5194/gmd-10-3679-2017

Special issue: The Modular Earth Submodel System (MESSy) (ACP/GMD inter-journal...

Geosci. Model Dev., 10, 3679–3693, 2017
https://doi.org/10.5194/gmd-10-3679-2017

Development and technical paper 10 Oct 2017

Development and technical paper | 10 Oct 2017

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

Michail Alvanos and Theodoros Christoudias Michail Alvanos and Theodoros Christoudias
  • The Cyprus Institute, P.O. Box 27456, 1645 Nicosia, Cyprus

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 CUDA-enabled 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.

Download
Short summary
We present an application of GPU accelerators in Earth system modeling. We developed software that generates CUDA kernels for numerical integration in the global climate model EMAC, used to study climate change and air quality. We focus on atmospheric chemical kinetics, the most computationally intensive task in climate–chemistry simulations. This approach can serve as the basis for hardware acceleration of numerous geoscientific models that rely on KPP for chemical kinetics applications.