1. FPGA Port of a Large Scientific Model from Legacy Code: The Emanuel Convection Scheme

Mr Kristian Hentschel, University of Glasgow

The potential of FPGAs for High-Performance Computing is increasingly being recognised, but most work focuses on acceleration of small, isolated kernels. We present a parallel FPGA implementation of a scienti c legacy algorithm, the seminal scheme for cumulus convection in large-scale models developed by Emanuel. Our design makes use of pipelines both at the arithmetic and at the logical stage level, keeping the entire algorithm on the FPGA. We assert that modern FPGAs have the resources to support this type of large algorithms. Through a practical and theoretical evaluation of our design we show how such an FPGA implementation compares to a CPU implementations or multi-core approaches such as OpenMP.

--------------------------------------------

2. Exploring the parallelisation of the Large Eddy Simulation using MPI and the Glasgow Model Coupling Framework

Mr Gordon Reid, University of Glasgow

Climate scientists write a number of applications to simulate aspects of our planet's ecosystem. These are sometimes coupled with other simulations using model coupling frameworks. This improves the accuracy of results and allows climate scientists to model more complicated scenarios. The Glasgow Model Coupling Framework (GMCF) is a new kind of model coupling framework that can also be used to parallelise individual models. By comparing LES performance when parallelised by MPI and GMCF, it is shown that GMCF offers competitive performance with MPI while also providing model coupling functionality.