Abstract. We have optimised the atmospheric radiation algorithm of the FAMOUS climate model on several hardware platforms. The optimisation involved translating the Fortran code to C and restructuring the algorithm around the computation of a single air column. Instead of the existing MPI-based domain decomposition, we used a task queue and a thread pool to schedule the computation of individual columns on the available processors. Finally, four air columns are packed together in a single data structure and computed simultaneously using Single Instruction Multiple Data operations.The modified algorithm runs more than 50 times faster on the CELL's Synergistic Processing Elements than on its main PowerPC processing element. On Intel-compatible processors, the new radiation code runs 4 times faster. On the tested graphics processor, using OpenCL, we find a speed-up of more than 2.5 times as compared to the original code on the main CPU. Because the radiation code takes more than 60 % of the total CPU time, FAMOUS executes more than twice as fast. Our version of the algorithm returns bit-wise identical results, which demonstrates the robustness of our approach. We estimate that this project required around two and a half man-years of work.

No abstract

We address the problem of metadata management in the context of future Electronic Music Distribution (EMD) systems, and propose a classification of existing musical editorial systems in two categories: the isolationists and the universalists. Universalists propose shared information at the expense of consensuality, while isolationist approaches allow individual parameterization at the expense of the lack of reusability. We propose an architecture and a system for managing editorial metadata that lies in the middle of these two extremes: we organize musical editorial information in such a way that users can benefit from shared metadata when they wish, while also allowing them to create and manage a private version of editorial information. A mechanism allows the synchronizing of both views: the shared and the private.

No abstract