Turbulence Start-up Option in CFD-ACE+ and CFD-CAD
作者:admin 日期:2008-08-26
Turbulence Start-up Option in CFD-ACE+ and CFD-CADalyzer
Turbulent flow can be hard to stabilize during the first iterations. Fortunately, a solution control feature exists in CFD-ACE+ and CADalyzer that can aid the convergence of turbulent cases. Consider the case study presented in figure 1.
Figure 1: Example case study
This is a small circular jet entering into a rectangular shaped duct. The inlet conditions for the duct are 30.0 m/s with a turbulence intensity of 0.03 and for the jet are 71.263 m/s with a turbulence intensity of 0.05. The initial conditions for this case are 30.0 m/s for U velocity, 0.1 for turbulence intensity, and 0.203 for the dissipation rate. Using default values for relaxation the simulation will actually diverge due to poor initial conditions. However, using the new Turbulent Start Control feature will allow the case to converge. The residual plots from each of the two cases are shown in figures 2 and 3.
Figure 2: Residuals without turbulence start-up option
Turbulent flow can be hard to stabilize during the first iterations. Fortunately, a solution control feature exists in CFD-ACE+ and CADalyzer that can aid the convergence of turbulent cases. Consider the case study presented in figure 1.
Figure 1: Example case study
This is a small circular jet entering into a rectangular shaped duct. The inlet conditions for the duct are 30.0 m/s with a turbulence intensity of 0.03 and for the jet are 71.263 m/s with a turbulence intensity of 0.05. The initial conditions for this case are 30.0 m/s for U velocity, 0.1 for turbulence intensity, and 0.203 for the dissipation rate. Using default values for relaxation the simulation will actually diverge due to poor initial conditions. However, using the new Turbulent Start Control feature will allow the case to converge. The residual plots from each of the two cases are shown in figures 2 and 3.
Figure 2: Residuals without turbulence start-up option
Option to Not Write the Restart Solution Data to D
作者:admin 日期:2008-08-20
Option to Not Write the Restart Solution Data to DTF Files
As simulation file sizes get bigger and bigger, we are always looking for ways to decrease the amount of data stored. A new option has been added to CFD-ACE+ that allows you to NOT save the restart data to the DTF file. This can reduce file size by as much as 40% for some large models.
Basically, the DTF file has two sets of solution data: one is for visualization (what you see in CFD-VIEW) and one for restart purposes (allowing you to restart the simulation to run further or restart it with different parameters).
Many times, a simulation will be run only one time to convergence, and will never be used as a restart point. Up until now, both the visualization and restart data were written into DTF files, which can make them unnecessarily large.
A brand new option, now available in CFD-ACE+ version 2008.2.9, is the ability to choose whether or not the restart data will be written to the DTF file. This option can be found in the 'OUT' panel, under the 'Ouput' tab, as shown in figure 1.
Figure 1: Restart data check box in CFD-ACE-GUI
By default, the "Write Restart Solution Data" option is activated, meaning that the restart data WILL be written to the DTF file. Uncheck it and the restart data will not be written to the DTF file, thus making it much smaller, up to 40% size reduction. Be careful though, as you will not be able to use the file as initial conditions for steady or unsteady simulations.
If you have any questions about this feature or would like us to discuss some other topic in the future, please let us know.
As simulation file sizes get bigger and bigger, we are always looking for ways to decrease the amount of data stored. A new option has been added to CFD-ACE+ that allows you to NOT save the restart data to the DTF file. This can reduce file size by as much as 40% for some large models.
Basically, the DTF file has two sets of solution data: one is for visualization (what you see in CFD-VIEW) and one for restart purposes (allowing you to restart the simulation to run further or restart it with different parameters).
Many times, a simulation will be run only one time to convergence, and will never be used as a restart point. Up until now, both the visualization and restart data were written into DTF files, which can make them unnecessarily large.
A brand new option, now available in CFD-ACE+ version 2008.2.9, is the ability to choose whether or not the restart data will be written to the DTF file. This option can be found in the 'OUT' panel, under the 'Ouput' tab, as shown in figure 1.
Figure 1: Restart data check box in CFD-ACE-GUI
By default, the "Write Restart Solution Data" option is activated, meaning that the restart data WILL be written to the DTF file. Uncheck it and the restart data will not be written to the DTF file, thus making it much smaller, up to 40% size reduction. Be careful though, as you will not be able to use the file as initial conditions for steady or unsteady simulations.
If you have any questions about this feature or would like us to discuss some other topic in the future, please let us know.
FE-DESIGN releases Version 6.2 of the Optimization
作者:admin 日期:2008-08-17
Press Release
FE-DESIGN releases Version 6.2 of the Optimization Software TOSCA Structure
Strongly enhanced functions of the software solution for efficient structural optimization
Karlsruhe, 15. August 2008
FE-DESIGN is enhancing the reliable approach of solver independent structural optimization in combination with the commercial finite element solvers ABAQUS, ANSYS, MSC.Nastran, MSC.Marc, NX.Nastran and PERMAS in the area if nonlinear finite element analysis.
In Topology Optimization, several non-linear constitutive materials are allowed to be a part of the design domain. This may from now on be combined with the definition of complex objectives and constraints. In Version 6.2, acoustic measures may be used for topology optimization. For the formulation of objectives and constraints, the surface velocities as well as the pressure level may be used. The support of internal forces allows the application of contact forces or forces of weld elements in the objective or constraints. In the area of manufacturing restrictions, the rib distance was implemented for the optimization of reinforcement ribs for casting parts.
Also in Shape Optimization, enhancements in the area of non-linear analysis have been made. In TOSCA Structure Version 6.2, the plastic strains are allowed to be used as design response for the objective formulation. For the manufacturing constraints, a new sliding constraint is implemented that allows the movement of the surface nodes on arbitrary free form surfaces.
In Bead Optimization, the proven sensitivity based algorithm is implemented that allows the definition of complex optimization tasks for the optimization of sheet metal components. In Version 6.2 compliance, displacements, modal frequencies and results from frequency response analysis may be used for bead optimization.
The smoothing algorithms in TOSCA Structure.smooth were redesigned and the detection and avoidance of self intersections was added. The definition of the smoothing task may now be added to the parameter file of TOSCA and therefore be executed automatically at the end of the optimization run.
Tosca Structure 6.2新版本功能
作者:admin 日期:2008-08-12
We are pleased to present you the latest version of TOSCA Structure 6.2, containing a large number of new functionalities:
Here some highlights:
Topology Optimization:
- Internal forces to be used as design response for objectives and constraints
- Design Responses from acoustics and frequency response analysis
- Definition of multiple volume constraints possible
- Support of geometric nonlinear analysis and contact in the design area for ABAQUS and ANSYS
- Support of a number of nonlinear materials in combination with ABAQUS
- Soft Delete
- Definition of rib distances for wall thickness constraints (maximum membersize).
Shape Optimization
- Support of plastic strains as design response
- Check Slide for manufacturing
- Support of CPRESS for the contact surface shape optimization with ABAQUS
Bead Optimization
Here some highlights:
Topology Optimization:
- Internal forces to be used as design response for objectives and constraints
- Design Responses from acoustics and frequency response analysis
- Definition of multiple volume constraints possible
- Support of geometric nonlinear analysis and contact in the design area for ABAQUS and ANSYS
- Support of a number of nonlinear materials in combination with ABAQUS
- Soft Delete
- Definition of rib distances for wall thickness constraints (maximum membersize).
Shape Optimization
- Support of plastic strains as design response
- Check Slide for manufacturing
- Support of CPRESS for the contact surface shape optimization with ABAQUS
Bead Optimization
Mixing Structured and Unstructured 2D Grids
作者:admin 日期:2008-08-12
Mixing Structured and Unstructured 2D Grids
One of the most common questions that occurs during CFD-GEOM Training sessions is "How do I generate mixed structured and unstructured grids?" Both grid types have their specific advantages and disadvantages, e.g. structured grids are very efficient in resolving the viscous terms, whereas unstructured grids can be used to handle complex geometries.
This example demonstrates complex geometry grids for a 2-dimensional model simulating a microfluidic DNA chamber containing five pillars. Using the following steps, node to node matching at the interface between the structured and unstructured domain is achieved, and a grid such as the one shown below can be created.
1. Create the Geometry Out of Points, Lines and Curves:
Your computational area of interest is now shown as a wire frame.
2. Create the Structured Domain:
Create edges on the lines/curves with the desired number of points and point distribution. Once edge creation is complete, create a structured face and 2D block for the channel.
3. Create Trimmed Surfaces:
One of the most common questions that occurs during CFD-GEOM Training sessions is "How do I generate mixed structured and unstructured grids?" Both grid types have their specific advantages and disadvantages, e.g. structured grids are very efficient in resolving the viscous terms, whereas unstructured grids can be used to handle complex geometries.
This example demonstrates complex geometry grids for a 2-dimensional model simulating a microfluidic DNA chamber containing five pillars. Using the following steps, node to node matching at the interface between the structured and unstructured domain is achieved, and a grid such as the one shown below can be created.
1. Create the Geometry Out of Points, Lines and Curves:
Your computational area of interest is now shown as a wire frame.
2. Create the Structured Domain:
Create edges on the lines/curves with the desired number of points and point distribution. Once edge creation is complete, create a structured face and 2D block for the channel.
3. Create Trimmed Surfaces:
Plotter Operations with CFD-VIEW
作者:admin 日期:2008-08-02
Plotter Operations with CFD-VIEW
The Plotter operator in CFD-VIEW supports signal processing for Time History data. A previous user tip – Digital Signal Processing using CFD-VIEW – shows how time history date can be made available and how to perform a PSD on a periodic signal. It is recommended to review this user tip before going further.The aim of this new user tip is to list all options supported by the Plotter Operator.
This tip will discuss the following operations:
- Correlation
- Fast Fourier Transform
- Discrete Fourier Transform
- Integral
- Derivative
- Power Spectral Density
- Smoothing
- Filter
- Sound Pressure Level
- Window
If you are interested in these functions, please read on.
The Plotter operator in CFD-VIEW supports signal processing for Time History data. A previous user tip – Digital Signal Processing using CFD-VIEW – shows how time history date can be made available and how to perform a PSD on a periodic signal. It is recommended to review this user tip before going further.The aim of this new user tip is to list all options supported by the Plotter Operator.
This tip will discuss the following operations:
- Correlation
- Fast Fourier Transform
- Discrete Fourier Transform
- Integral
- Derivative
- Power Spectral Density
- Smoothing
- Filter
- Sound Pressure Level
- Window
If you are interested in these functions, please read on.
Restarting a Fine Grid Case from a Coarse Grid Sol
作者:admin 日期:2008-07-25
Restarting a Fine Grid Case from a Coarse Grid Solution
Many users have asked for a method to obtain a solution on a coarse grid and then use it as the initial guess for a finer grid. This option can be done using the Map utility in CFD-Toolkit.
The steps to perform this operation are as follows:
Create coarse grid system: The first step is to create the coarse grid system of the geometry in CFD-GEOM. In this example the following coarse grid is used. Save the coarse grid system as a DTF file, such as coarse_grid.DTF.
Set up and run the problem: Open the coarse_grid.DTF file in CFD-ACE-GUI and set up the problem. Once the setup is complete, submit it to the solver to be run.
Copy the Simulation: Once the simulation has completed running, copy the coarse_grid.DTF file into another file, such as fine_grid.DTF. The fine_grid.DTF has all the settings that the coarse_grid.DTF has, but still has the coarse grid system.
Create the Fine Grid: Go to CFD-GEOM and open the coarse_grid.GGD file if it is not already open. Create the fine grid now by changing the number of grid points. Only the number of grid points are changed and no modifications are made to the geometry shape. Also be careful not to delete or recreate any entity. Save the DTF file as fine_grid.DTF by selecting the Update Grids Only option. Now fine_grid.DTF consists of the finer grid system but retains all the model setup options from the coarse grid simulation.
Map Solution Data: The solution data in the fine_grid.DTF file is still from the coarse grid solution and therefore not consistent with the fine grid system. To map the solution data from the coarse grid DTF file to the fine grid DTF file, we can use CFD-Toolkit. The first thing to do is launch CFD-Toolkit. (Note: On Windows systems, choose Start-->Run-->CFD-Toolkit.) Go to the File View/Manipulate --> Map DTF File. Once it has been selected, the following window will be active.
Many users have asked for a method to obtain a solution on a coarse grid and then use it as the initial guess for a finer grid. This option can be done using the Map utility in CFD-Toolkit.
The steps to perform this operation are as follows:
Create coarse grid system: The first step is to create the coarse grid system of the geometry in CFD-GEOM. In this example the following coarse grid is used. Save the coarse grid system as a DTF file, such as coarse_grid.DTF.
Set up and run the problem: Open the coarse_grid.DTF file in CFD-ACE-GUI and set up the problem. Once the setup is complete, submit it to the solver to be run.
Copy the Simulation: Once the simulation has completed running, copy the coarse_grid.DTF file into another file, such as fine_grid.DTF. The fine_grid.DTF has all the settings that the coarse_grid.DTF has, but still has the coarse grid system.
Create the Fine Grid: Go to CFD-GEOM and open the coarse_grid.GGD file if it is not already open. Create the fine grid now by changing the number of grid points. Only the number of grid points are changed and no modifications are made to the geometry shape. Also be careful not to delete or recreate any entity. Save the DTF file as fine_grid.DTF by selecting the Update Grids Only option. Now fine_grid.DTF consists of the finer grid system but retains all the model setup options from the coarse grid simulation.
Map Solution Data: The solution data in the fine_grid.DTF file is still from the coarse grid solution and therefore not consistent with the fine grid system. To map the solution data from the coarse grid DTF file to the fine grid DTF file, we can use CFD-Toolkit. The first thing to do is launch CFD-Toolkit. (Note: On Windows systems, choose Start-->Run-->CFD-Toolkit.) Go to the File View/Manipulate --> Map DTF File. Once it has been selected, the following window will be active.
最新版本CFdesign v10发布
作者:admin 日期:2008-07-24
美国蓝脊数码有限公司于7月1日正式发布最新版本CFdesign v10。
CFdesign v10是CFdesign软件历史上的一次大版本升级。 V10使“Upfront CFD”的技术革新又向前迈进了一大步。它兼备现代化环境的实用性及众多出色的全新功能，成功地向设计师和工程师呈现精确到位的模拟工作流，是一套与众不同的 CFD 解决方案。
CFdesign v10 结合了具有前瞻性的创新技术和现今市场流行的解决方案。其中，可模拟空化和热电制冷器的各种新增功能，以及快速执行概念化设计迭代的“快速对流”模式，都是 v10 软件包的主要元素。 新增加的一套几何图形工具大大简化了制作复杂的三维装配几何的过程。CAD 组件 AutoPrep 已直接加到 CAD 环境中，模拟部件的几何创建功能也出现在 CFdesign 环境里。而这两者对于诊断、修复和准备用于分析的设计图形均十分关键。 通过在 CFdesign 环境中保持 CAD 主要元素的可用性，进一步深化了 CAD 与 CFdesign 模拟环境的联系。改进的用户界面使 CFdesign 工作流程更易掌握，更加有。
“CFdesign 分布式计算”概念的引入，让多方案设计同时进行成为可能。随着全球市场压力的不断增大，快速改善模拟结果对于创新、有效的设计至关重要。“CFdesign 分布式计算”可让使用“前端 CFD”技术的产品开发组织充分利用先进的高速网络计算环境，在激烈的竞争中脱颖而出。
请点击我们的官方网站以获取有关新版本的更多信息： caeda@vip.sina.com
CFdesign v10是CFdesign软件历史上的一次大版本升级。 V10使“Upfront CFD”的技术革新又向前迈进了一大步。它兼备现代化环境的实用性及众多出色的全新功能，成功地向设计师和工程师呈现精确到位的模拟工作流，是一套与众不同的 CFD 解决方案。
CFdesign v10 结合了具有前瞻性的创新技术和现今市场流行的解决方案。其中，可模拟空化和热电制冷器的各种新增功能，以及快速执行概念化设计迭代的“快速对流”模式，都是 v10 软件包的主要元素。 新增加的一套几何图形工具大大简化了制作复杂的三维装配几何的过程。CAD 组件 AutoPrep 已直接加到 CAD 环境中，模拟部件的几何创建功能也出现在 CFdesign 环境里。而这两者对于诊断、修复和准备用于分析的设计图形均十分关键。 通过在 CFdesign 环境中保持 CAD 主要元素的可用性，进一步深化了 CAD 与 CFdesign 模拟环境的联系。改进的用户界面使 CFdesign 工作流程更易掌握，更加有。
“CFdesign 分布式计算”概念的引入，让多方案设计同时进行成为可能。随着全球市场压力的不断增大，快速改善模拟结果对于创新、有效的设计至关重要。“CFdesign 分布式计算”可让使用“前端 CFD”技术的产品开发组织充分利用先进的高速网络计算环境，在激烈的竞争中脱颖而出。
请点击我们的官方网站以获取有关新版本的更多信息： caeda@vip.sina.com
基于Windows的高性能计算解决方案
作者:lns 日期:2008-07-23
Microsoft 在多年之前就开始支持容错和高可用性集群，而随着 Windows Compute Cluster Server 2003 的发布，Microsoft 将高性能计算 (HPC) 的超级计算能力引入了个人和工作组领域。 从2006年夏季发布WCCS 2003至今, Microsoft WCCS 由于其拥有的简单易用等特点而引起了业界的广泛关注。在2008年6月超级计算大会公布的全球高性能计算TOP500中，有5套采用Windows的超级计算系统，在前100强中就有3套，其中NCSA排在23位，Linpack测试的运行效率达到了76.4%，排在39位的Umea University 是一套在IBM 刀片服务器HS21运行上Windows的高性能计算平台，运行效率更是高达85.64% ，该系统也是在TOP500中Linpack的运行效率是最高的几套集群系统之一。
Microsoft作为软件行业的领军企业，长期以来致力于帮助各类用户解决在日常办公和科研/生产/管理中的遇到的计算机应用问题，开发和发展Microsoft的HPC软件也是为了更好地服务于用户，让更多的用户透过微软提供的简单易用的高性能计算平台告别单机处理计算的时代，采用多机并行的方式来大大提高工作效率，缩短科研生产周期，进而提升企业的竞争力。
Microsoft的高性能计算软件WCCS是建构在PC集群基础上的，除操作系统外，还集成了功能完善的集群系统管理、监控、MSMPI、作业调度、日志报告等功能，与传统的Linux集群系统相比较，更容易部署和管理。WCCS 2003的基本组件有两张CD：第一张 CD 包含 Windows Server 2003 Compute Cluster Edition， 第二张CD是Microsoft Compute Cluster Pack — 构成 Windows Compute Cluster Server 2003的界面、实用工具和管理基础架构的组合。
Microsoft WCCS 2003 提供了一个运行在商用 x64 计算机集群上的极经济的强大 HPC 解决方案，可使用熟悉的工具和技术，轻松、快速地对其进行扩展和部署。其基本体系架构描述如下：包含单个头节点（也叫主控节点）和多个计算节点（参见图 1）。头节点控制和协调所有对集群资源的访问，实现计算集群的单一管理、部署和作业调度。WCCS 2003 使用 Microsoft Operations Manager 2005 和 Microsoft Systems Management Server 2003 等工具，将现有的公司 Active Directory 基础结构用于安全、帐户管理和总体操作管理。
图 1. 典型的 WCCS 2003 网络
WCCS 2003的安装包括在头节点上安装操作系统，将头节点加入现有的 Active Directory 域，然后安装 Compute Cluster Pack。当 Compute Cluster Pack 安装完成后，将会弹出一个“任务列表”页面，其中显示完成计算集群配置所需执行的步骤。这些步骤中包括定义网络拓扑，使用“配置 RIS”向导配置 RIS，将计算节点添加到集群，以及配置集群用户和管理员等。 通过安装向导，可以指导用户方便地一步一步完成整个集群系统的部署。
微软的WCCS 2003 虽然涉足高性能计算时间较短，但其继承了微软视窗系统的图形化的管理风格，具有部署简单，管理容易，功能齐全，能有效地与企业已有Windows架构的IT系统方便集成，并且有更加简单熟悉的Visual Studio集成开发环境帮助技术人员快速进行并行程序开发设计。因为门槛低，用户可以快速上手，因此微软的WCCS 2003 越来越广泛地应用在工程设计/科学计算/图形图像/金融分析等高性能计算领域。
具体来讲，WCCS的优势集中体现在如下几个方面：
灵活的扩展性可充分保护用户的投资
Microsoft作为软件行业的领军企业，长期以来致力于帮助各类用户解决在日常办公和科研/生产/管理中的遇到的计算机应用问题，开发和发展Microsoft的HPC软件也是为了更好地服务于用户，让更多的用户透过微软提供的简单易用的高性能计算平台告别单机处理计算的时代，采用多机并行的方式来大大提高工作效率，缩短科研生产周期，进而提升企业的竞争力。
Microsoft的高性能计算软件WCCS是建构在PC集群基础上的，除操作系统外，还集成了功能完善的集群系统管理、监控、MSMPI、作业调度、日志报告等功能，与传统的Linux集群系统相比较，更容易部署和管理。WCCS 2003的基本组件有两张CD：第一张 CD 包含 Windows Server 2003 Compute Cluster Edition， 第二张CD是Microsoft Compute Cluster Pack — 构成 Windows Compute Cluster Server 2003的界面、实用工具和管理基础架构的组合。
Microsoft WCCS 2003 提供了一个运行在商用 x64 计算机集群上的极经济的强大 HPC 解决方案，可使用熟悉的工具和技术，轻松、快速地对其进行扩展和部署。其基本体系架构描述如下：包含单个头节点（也叫主控节点）和多个计算节点（参见图 1）。头节点控制和协调所有对集群资源的访问，实现计算集群的单一管理、部署和作业调度。WCCS 2003 使用 Microsoft Operations Manager 2005 和 Microsoft Systems Management Server 2003 等工具，将现有的公司 Active Directory 基础结构用于安全、帐户管理和总体操作管理。
图 1. 典型的 WCCS 2003 网络
WCCS 2003的安装包括在头节点上安装操作系统，将头节点加入现有的 Active Directory 域，然后安装 Compute Cluster Pack。当 Compute Cluster Pack 安装完成后，将会弹出一个“任务列表”页面，其中显示完成计算集群配置所需执行的步骤。这些步骤中包括定义网络拓扑，使用“配置 RIS”向导配置 RIS，将计算节点添加到集群，以及配置集群用户和管理员等。 通过安装向导，可以指导用户方便地一步一步完成整个集群系统的部署。
微软的WCCS 2003 虽然涉足高性能计算时间较短，但其继承了微软视窗系统的图形化的管理风格，具有部署简单，管理容易，功能齐全，能有效地与企业已有Windows架构的IT系统方便集成，并且有更加简单熟悉的Visual Studio集成开发环境帮助技术人员快速进行并行程序开发设计。因为门槛低，用户可以快速上手，因此微软的WCCS 2003 越来越广泛地应用在工程设计/科学计算/图形图像/金融分析等高性能计算领域。
具体来讲，WCCS的优势集中体现在如下几个方面：
灵活的扩展性可充分保护用户的投资
学 术 报 告
作者:lns 日期:2008-07-22
学 术 报 告
报告题目（一）:
Fault Tolerance for PetaScale Systems: Current Knowledge, Challenges and Opportunities
报告人：Professor Franck Cappello( INRIA)
时间：2008年7月25日（星期五）上午9：30-10：30
地点：软件所5号楼3层337会议室
简介: Professor Franck Cappello holds a Senior Researcher position at INRIA. He leads the Grand-Large project at INRIA, focusing on High Performance issues in Large Scale Distributed Systems. He has initiated the XtremWeb (Desktop Grid) and MPICH-V (Fault tolerant MPI) projects. He is currently the director of the Grid5000 project, a nation wide computer science platform for research in Grid and P2P. He has authored more than 60 papers in the domains of High Performance Programming, Desktop Grids, Grids and Fault tolerant MPI. He has contributed to more than 40 Program Committees. He is editorial board member of the international Journal on Grid Computing, Journal of Grid and Utility Computing and Journal of Cluster Computing.He is a steering committee member of IEEE HPDC and IEEE/ACM CCGRID.He is the General co-Chair of IEEE APSCC 2008, Workshop co-chair for IEEE CCGRID'2008, Program co-Chair of IEEE CCGRID'2009 and was the General Chair of IEEE HPDC'2006.
报告摘要:
The emergence of PetaScale systems reinvigorates the community interest about how to manage failures in such systems and ensure that large applications successfully complete. Existing results for several key mechanisms associated with fault tolerance in HPC platforms will be presented during this talk. Most of these key mechanisms come from the distributed system theory. Over the last decade, they have received a lot of attention from the community and there is probably little to gain by trying to optimize them again. We will describe some of the latest findings in this domain. Unfortunately, despite their high degree of optimization, existing approaches do not fit well with the challenging evolutions of large scale systems. There is room and even a need for new approaches. Opportunities may come from different origins like adding hardware dedicated to fault tolerance. We will sketch some of these opportunities and their associated limitations
报告题目（一）:
Fault Tolerance for PetaScale Systems: Current Knowledge, Challenges and Opportunities
报告人：Professor Franck Cappello( INRIA)
时间：2008年7月25日（星期五）上午9：30-10：30
地点：软件所5号楼3层337会议室
简介: Professor Franck Cappello holds a Senior Researcher position at INRIA. He leads the Grand-Large project at INRIA, focusing on High Performance issues in Large Scale Distributed Systems. He has initiated the XtremWeb (Desktop Grid) and MPICH-V (Fault tolerant MPI) projects. He is currently the director of the Grid5000 project, a nation wide computer science platform for research in Grid and P2P. He has authored more than 60 papers in the domains of High Performance Programming, Desktop Grids, Grids and Fault tolerant MPI. He has contributed to more than 40 Program Committees. He is editorial board member of the international Journal on Grid Computing, Journal of Grid and Utility Computing and Journal of Cluster Computing.He is a steering committee member of IEEE HPDC and IEEE/ACM CCGRID.He is the General co-Chair of IEEE APSCC 2008, Workshop co-chair for IEEE CCGRID'2008, Program co-Chair of IEEE CCGRID'2009 and was the General Chair of IEEE HPDC'2006.
报告摘要:
The emergence of PetaScale systems reinvigorates the community interest about how to manage failures in such systems and ensure that large applications successfully complete. Existing results for several key mechanisms associated with fault tolerance in HPC platforms will be presented during this talk. Most of these key mechanisms come from the distributed system theory. Over the last decade, they have received a lot of attention from the community and there is probably little to gain by trying to optimize them again. We will describe some of the latest findings in this domain. Unfortunately, despite their high degree of optimization, existing approaches do not fit well with the challenging evolutions of large scale systems. There is room and even a need for new approaches. Opportunities may come from different origins like adding hardware dedicated to fault tolerance. We will sketch some of these opportunities and their associated limitations
Setting the Viewer Size using Scripting in CFD-VIE
作者:admin 日期:2008-07-19
Setting the Viewer Size using Scripting in CFD-VIEW
When post-processing results, you often want to compare two different models in the exact same position and size. The ‘Copy and Past Viewpoint’ feature allows you to set the same position and angle to display your model but, in addition to that, you might also want to have the exact same size for the two windows you are comparing. This would ensure that all the pictures in your report have the same original dimensions.
One new Python command will let you do that very easily.
This command is:
SetViewerSize(w,h)
where w is the width of the active viewer and h is the height.
To use it, you can click on the ‘Edit Python script’ icon:
Figure 1: 'Edit Python script' icon
When post-processing results, you often want to compare two different models in the exact same position and size. The ‘Copy and Past Viewpoint’ feature allows you to set the same position and angle to display your model but, in addition to that, you might also want to have the exact same size for the two windows you are comparing. This would ensure that all the pictures in your report have the same original dimensions.
One new Python command will let you do that very easily.
This command is:
SetViewerSize(w,h)
where w is the width of the active viewer and h is the height.
To use it, you can click on the ‘Edit Python script’ icon:
Figure 1: 'Edit Python script' icon
2008@西安 高性能计算培训通知
作者:lns 日期:2008-07-15
2008@西安 高性能计算培训通知
为培养高性能计算专业人才、推广高性能计算应用、加强与国家科研单位、高等院校及企业的交流与合作，推动区域高性能计算应用与发展，中国科学院计算机网络信息中心超级计算中心在中科院西安分院的大力支持下，将于2008年7月28-29日在西安举办高性能计算培训班。
具体通知如下：
l 培训对象：具备高性能计算和Linux基础知识，已经从事与高性能计算应用相关的科研院所及高校老师、硕博研究生及相关企业工程师，或对高性能计算感兴趣的相关人员。
l 培训时间：2008年7月28日－7月29日
l 培训地点：西安，东方大酒店（西安市小寨西路子午路）
住宿费：200元/标间/天、210元/单间/天
为培养高性能计算专业人才、推广高性能计算应用、加强与国家科研单位、高等院校及企业的交流与合作，推动区域高性能计算应用与发展，中国科学院计算机网络信息中心超级计算中心在中科院西安分院的大力支持下，将于2008年7月28-29日在西安举办高性能计算培训班。
具体通知如下：
l 培训对象：具备高性能计算和Linux基础知识，已经从事与高性能计算应用相关的科研院所及高校老师、硕博研究生及相关企业工程师，或对高性能计算感兴趣的相关人员。
l 培训时间：2008年7月28日－7月29日
l 培训地点：西安，东方大酒店（西安市小寨西路子午路）
住宿费：200元/标间/天、210元/单间/天
