Dr. Xiaohu Guo,Senior Computational Scientist, Computational Science&Engineering,Department, Science and Technology Facilities Council, Daresbury Laboratory,Warrington WA4 4AD, United Kingdom

High Performance Computing Driven Software Development for Next-Generation Modeling of the Worlds Oceans

The Imperial College Ocean Model (ICOM) is an open-source next generation ocean model build upon finite element methods and anisotropic unstructured adaptive meshing. Since 2009, a project has been funded by EPSRC to optimise the ICOM for the UK national high-end computing resource, HECToR. Major developments and optimizations have been carried out to address the performance issues of unstructured mesh models, such as indirect addressing and cache reuse(a particularly severe challenge for unstructured mesh methods). Various profiling tools such as CrayPAT and Vampir, have been extensively used to analysis issues of the scalability of matrix assembly,the sparse linear solvers/preconditioners. Hybrid(OpenMP/MPI) parallelisms have also been explored in order to further scale Fluidity-ICOM onto petascale platforms. Presently the code is scaling well up to 4096 cores on HECToR. Runs on even larger core counts could be achieved if suitably partitioned datasets existed. There will be a discussion about the current fluidity-ICOM software development strategy.