B. Min, S. Suckow, U. Yusufoglu, T. Pletzer, H. Kurz
{"title":"Acceleration of 3D numerical simulation of silicon solar cell using thread parallelism","authors":"B. Min, S. Suckow, U. Yusufoglu, T. Pletzer, H. Kurz","doi":"10.1109/NUSOD.2012.6316540","DOIUrl":null,"url":null,"abstract":"We have investigated the potential to accelerate three-dimensional numeric simulation of silicon solar cell using thread parallelism. The device simulated is a rear side passivated cell with rear point contacts (PERC). The optical and electrical behaviour of the device was simulated with Sentaurus Device (formerly dessis). We show that the simulation run-time on a four socket Opteron 6168 machine is reduced down to 6% compared to the run-time without thread parallelism. Furthermore, limits of time reduction by varying the number of threads up to 48 are studied. Thereby, the number of threads for the optimum use of the hardware resources is determined.","PeriodicalId":337826,"journal":{"name":"2012 12th International Conference on Numerical Simulation of Optoelectronic Devices (NUSOD)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 12th International Conference on Numerical Simulation of Optoelectronic Devices (NUSOD)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/NUSOD.2012.6316540","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
We have investigated the potential to accelerate three-dimensional numeric simulation of silicon solar cell using thread parallelism. The device simulated is a rear side passivated cell with rear point contacts (PERC). The optical and electrical behaviour of the device was simulated with Sentaurus Device (formerly dessis). We show that the simulation run-time on a four socket Opteron 6168 machine is reduced down to 6% compared to the run-time without thread parallelism. Furthermore, limits of time reduction by varying the number of threads up to 48 are studied. Thereby, the number of threads for the optimum use of the hardware resources is determined.