Global optimization algorithms for a cad workstation
Journal of Optimization Theory and Applications
Power programming with RPC
Euro-Par '99 Proceedings of the 5th International Euro-Par Conference on Parallel Processing
WSEAS Transactions on Computers
| Hi-index | 0.00 |
Computer simulations are becoming a fundamental instrument in order to determine the suffer effects of a specific model by varying its input parameters. However, to obtain simulation results, the needed computational power required, by increasing the model parameters, in a reasonable time, is not so indifferent. So the aim of this research is to optimize computational throughput. At ST Microelectronics, a new integrated computational solution to resolve this problem in an efficient way has been implemented and tested. This solution consist of two blocks: a parallel implementation approach, named Exemplar, and a workload balancing system into a Cluster Grid. The first block was designed on a heterogeneous hardware platform which could have different computational performances. It is based on different optimization methods: CRS (Controlled Random Search) method, and Monte Carlo method. The second block consists of a cluster of CPU servers integrated through a workload balance system and a data resource management system (DRMS).