首页> 外文会议>High Performance Computing Symposium >Modeling tidal flow currents in the Taiwan strait on parallel computers
【24h】

Modeling tidal flow currents in the Taiwan strait on parallel computers

机译:在平行电脑台湾海峡拓展潮流电流

获取原文

摘要

Tidal movement is the dominant force causing shelf sea circulation. Variation of the climate creates seasonal circulation. A high performance shelf sea hydrodynamic model can enhance the understanding of various marine processes. The shelf sea area around Taiwan has complicated features mainly due to the interaction with the Pacific Ocean current. Pollution problems associated with high-density exploitation of the marine resources, coastal constructions and marine disposal facilities. A large-area fine resolution hydrodynamic model can provide long term and detailed hydrodynamic information for the assessment of various coastal and marine developments and constructions. It can provide information for marine pollution simulation and for hazardous release management. Grid size becomes small when detailed local resolution is required. In order to maintain the same model boundary features, computations become enormous. Conventional sequential model cannot efficiently provide simulation results. In this paper, a tidal flow model, which covers an area of the Taiwan Strait and its adjacent seas, has been implemented on several parallel machines. These contain either shared memory machines and distributed memory machines. Since PC-clusters become popular, the experiences of porting the model on PC-clusters will be included in this paper. The performances among various types of parallel computers will also be presented and discussed.
机译:潮汐运动是导致货架海洋循环的主导力。气候的变异会产生季节性循环。高性能架子水动力模型可以增强对各种海洋过程的理解。台湾周围的货架海域具有复杂的功能,主要是由于与太平洋电流的互动。与海洋资源,沿海建筑和海洋处置设施高密度开采相关的污染问题。大面积的精细分辨率流体动力学模型可以提供长期和详细的流体动力学信息,用于评估各种沿海和海洋发展和建筑。它可以为海洋污染模拟和危险释放管理提供信息。当需要详细的本地分辨率时,网格尺寸变小。为了保持相同的模型边界特征,计算变得巨大。传统的顺序模型无法有效地提供仿真结果。本文在几台平行机上实现了覆盖了台湾海峡地区及其相邻海洋的潮流模型。这些包含共享内存计算机和分布式存储器。由于PC-Clusters变得流行,因此本文将包含在PC集群上移植模型的经验。还将呈现和讨论各种类型的并行计算机之间的性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号