Ongoing climate change calls for fast and accurate weather and climate modeling. However, when solving large-scale weather prediction simulations, state-of-the-art CPU and GPU implementations suffer from limited performance and high energy consumption. These implementations are dominated by complex irregular memory access patterns and low arithmetic intensity that pose fundamental challenges to acceleration. To overcome these challenges, we propose and evaluate the use of near-memory acceleration using a reconfigurable fabric with high-bandwidth memory (HBM). We focus on compound stencils that are fundamental kernels in weather prediction models. By using high-level synthesis techniques, we develop NERO, an FPGA+HBM-based accelerator connected through IBM CAPI2 (Coherent Accelerator Processor Interface) to an IBM POWER9 host system. Our experimental results show that NERO outperforms a 16-core POWER9 system by 4.2x and 8.3x when running two different compound stencil kernels. NERO reduces the energy consumption by 22x and 29x for the same two kernels over the POWER9 system with an energy efficiency of 1.5 GFLOPS/Watt and 17.3 GFLOPS/Watt. We conclude that employing near-memory acceleration solutions for weather prediction modeling is promising as a means to achieve both high performance and high energy efficiency.
|Title of host publication
|2020 30th International Conference on Field-Programmable Logic and Applications (FPL)
|Nele Mentens, Leonel Sousa, Pedro Trancoso, Miquel Pericas, Ioannis Sourdis
|Institute of Electrical and Electronics Engineers
|Number of pages
|Published - 13 Oct 2020
|30th International Conference on Field Programmable Logic and Applications, FPL 2020 - Virtual, Gothenburg, Sweden
Duration: 31 Aug 2020 → 4 Sept 2020
|30th International Conference on Field Programmable Logic and Applications, FPL 2020
|31/08/20 → 4/09/20
This work was performed in the framework of the Horizon 2020 program for the project “Near-Memory Computing (NeMeCo)”. It is funded by the European Commission under Marie Sklodowska-Curie Innovative Training Networks European Industrial Doctorate (Project ID: 676240). Special thanks to Florian Auernhammer and Raphael Polig for providing access to IBM systems. We appreciate valuable discussions with Kaan Kara and Ronald Luijten. We would also like to thank Bruno Mesnet and Alexandre Castellane from IBM France for help with the SNAP framework. This work was partially supported by the H2020 research and innovation programme under grant agreement No 732631, project OPRE-COMP. We also thank Google, Huawei, Intel, Microsoft, SRC, and VMware for their funding support.
|European Union’s H2020 research and innovation programme
|European Union 's Horizon 2020 - Research and Innovation Framework Programme