Supercomput. Front. Innov.最新文献

Multistage Iterative Method to Tackle Inverse Problems of Wave Tomography 多阶段迭代法求解波层析成像反演问题

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220106

A. Goncharsky, S. Romanov, S. Seryozhnikov

引用次数: 0

4D Technology of Variational Data Assimilation for Sea Dynamics Problems 海洋动力问题变分数据同化的4D技术

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220101

V. Shutyaev, V. Agoshkov, V. Zalesny, E. Parmuzin, N. Zakharova

引用次数: 0

River Routing in the INM RAS-MSU Land Surface Model: Numerical Scheme and Parallel Implementation on Hybrid Supercomputers INM RAS-MSU地表模型中的河流路径:数值方案与混合超级计算机上的并行实现

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220103

V. Stepanenko

{"title":"River Routing in the INM RAS-MSU Land Surface Model: Numerical Scheme and Parallel Implementation on Hybrid Supercomputers","authors":"V. Stepanenko","doi":"10.14529/jsfi220103","DOIUrl":"https://doi.org/10.14529/jsfi220103","url":null,"abstract":"The land surface model (LSM) is a necessary compartment of any numerical weather forecast system or the Earth system model. This paper presents a new version of the INM RAS-MSU land surface model where the river hydrodynamic and thermodynamic scheme is embedded into the parallel execution framework using MPI and OpenMP. Numerical experiments have been performed for the East European domain with resolution 0 . 5 ◦ × 0 . 5 ◦ . The soil model parallel eﬃciency at 1–144 MPI cores was 0.52–0.79 and limited by the presence of ocean area, and by imbalance of computational load between soil columns. The acceleration of the river model at MPI level was deﬁned by the size of the largest river basin in the domain. At the OpenMP level, the potential for acceleration of large river basin simulation is shown to be close to number of threads used, based on fractal properties of the river networks. This acceleration was hindered in our numerical experiments by the reduced river orders at the coarse land surface model resolution, so that the optimal speedup for the Volga river basin was 2.5–3 times attained at 4–6 threads. This performance is projected to improve with reﬁnement of the LSM spatial resolution. This paper presents a new version of the INM land where the river hydrodynamic and thermodynamic model is embedded into the parallel execution framework using two levels of parallelism: the ﬁrst is MPI-based indepedent processing of river basins, and the second uses OpenMP technique to parallelize the simulation of rivers of the same Strahler order. Numerical experiments have been performed for the East European domain with resolution 0 . 5 ◦ × 0 . 5 ◦ . The MPI implementation of the soil model is based on conventional even longitude-latitude decomposition of the model domain, inherited from the atmospheric model. The soil model parallel eﬃciency at 1–144 cores was shown to be 0.52–0.79 and limited by the presence of ocean area, and by imbalance of computational load between soil columns depending on the presence of snow cover and number of iterations for the surface temperature needed to advance the soil proﬁles. The acceleration of the river model at MPI level (not exceeding 4 times) is deﬁned by the size of the largest river basin in the domain (Volga), whereas at OpenMP level the potential for acceleration of large river basin simulation is shown to be close to number of threads used. OpenMP-level speedup was hindered in our numerical experiments by the underestimation of river orders at coarse land surface model resolution (recommended performance for the Volga basin attained at 4–6 threads with 2.5–3 times acceleration).","PeriodicalId":338883,"journal":{"name":"Supercomput. Front. Innov.","volume":"388 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116657524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Supercomputer-Based Modeling System for Short-Term Prediction of Urban Surface Air Quality 基于超级计算机的城市地表空气质量短期预报建模系统

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220102

A. Starchenko, E. Danilkin, Sergei A. Prokhanov, L. I. Kizhner, E. Shelmina

引用次数: 2

Machine Learning Approaches to Extreme Weather Events Forecast in Urban Areas: Challenges and Initial Results 城市地区极端天气事件预测的机器学习方法:挑战和初步结果

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220104

Fábio Porto, Mariza Ferro, Eduardo S. Ogasawara, T. Moeda, Claudio D. T. Barros, A. Silva, Rocío Zorrilla, R. S. Pereira, Rafaela Nascimento Castro, João Victor Silva, Rebecca Salles, Augusto Fonseca, Juliana Hermsdorff, Marcelo Magalhães, Vitor Sá, Adolfo Simões, Carlos Cardoso, Eduardo Bezerra

引用次数: 2

Data Assimilation by Neural Network for Ocean Circulation: Parallel Implementation 海洋环流数据的神经网络同化:并行实现

Supercomput. Front. Innov. Pub Date : 2022-03-01 DOI: 10.14529/jsfi220105

H. Velho, H. M. Furtado, S. B. Sambatti, Carla Barros Osthoff Ferreira de Barros, M. E. Welter, R. Souto, D. Carvalho, D. O. Cardoso

引用次数: 2

Direct Numerical Simulation of Stratified Turbulent Flows and Passive Tracer Transport on HPC Systems: Comparison of CPU Architectures 在高性能计算系统上分层湍流和被动示踪剂传输的直接数值模拟:CPU架构的比较

Supercomput. Front. Innov. Pub Date : 2021-12-01 DOI: 10.14529/jsfi210405

E. Mortikov, A. Debolskiy

引用次数: 2

High-performance Shallow Water Model for Use on Massively Parallel and Heterogeneous Computing Systems 用于大规模并行和异构计算系统的高性能浅水模型

Supercomput. Front. Innov. Pub Date : 2021-12-01 DOI: 10.14529/jsfi210407

A. Chaplygin, A. Gusev, N. Diansky