Method and system for checkpointing during simulations
    1.
    发明授权
    Method and system for checkpointing during simulations 有权
    模拟过程中检查点的方法和系统

    公开(公告)号:US08756042B2

    公开(公告)日:2014-06-17

    申请号:US13034341

    申请日:2011-02-24

    Abstract: Method and system for more efficient checkpointing strategy in cross correlating (316) a forward (328) and backward (308) propagated wave such as in migrating (326) or inverting seismic data. The checkpointing strategy includes storing in memory forward simulation data at a checkpointed time step, wherein the stored data are sufficient to do a cross correlation at that time step but not to restart the forward simulation. At other checkpoints, a greater amount of data sufficient to restart the simulation may be stored in memory (314). Methods are disclosed for finding an optimal combination, i.e. one that minimizes computation time (1132), of the two types of checkpoints for a given amount of computer memory (1004), and for locating a checkpoint at an optimal time step (306, 1214, 1310). The optimal checkpointing strategy (1002) also may optimize (1408) on use of fast (1402) vs. slow (1404) storage.

    Abstract translation: 交叉相关(316)前向(328)和向后(308)传播的波中的更有效的检查点策略的方法和系统,例如在迁移(326)或反转地震数据中。 检查点策略包括在检查点时间步骤中存储内存中的前向仿真数据,其中所存储的数据足以在该时间步骤进行互相关,但是不重新启动正向仿真。 在其他检查点,可以将足够重新启动模拟的更大量的数据存储在存储器(314)中。 公开了用于找到最佳组合的方法,即使给定量的计算机存储器(1004)的两种类型的检查点的计算时间(1132)最小化的最佳组合,并且用于在最佳时间步长(306,1214)处定位检查点 ,1310)。 最优检查点策略(1002)还可以优化(1408)快速(1402)与慢(1404)存储的使用。

    Method and system to reduce: memory requirements, device-to-host transfer bandwidth requirements, and setup time, for seismic modeling on graphics processing units
    2.
    发明授权
    Method and system to reduce: memory requirements, device-to-host transfer bandwidth requirements, and setup time, for seismic modeling on graphics processing units 有权
    减少图形处理单元地震建模的方法和系统,以减少内存要求,设备到主机传输带宽要求和建立时间

    公开(公告)号:US08972234B2

    公开(公告)日:2015-03-03

    申请号:US13367189

    申请日:2012-02-06

    CPC classification number: G01V1/28 G06F9/5044 G06T1/20 G06T15/005 G09G5/363

    Abstract: A method for seismic modeling on a GPU includes creating a discretized model containing a volume enclosed by boundaries with arbitrary boundary conditions, dividing the discretized model into subsets, transferring a data to the GPU, initializing the data stored in a selected number of the subsets on the GPU, performing computation related with a number of time steps for the selected number of the subsets on the GPU, and substantially simultaneously computing, off-loading a finished subsets data from the GPU, and up-loading a new subsets data to the GPU. The provided method and corresponding system can reduce the amount of data stored in GPU memory by an order of magnitude, minimize the GPU device-to-host transfer bandwidth requirements, and reduce the GPU setup time.

    Abstract translation: 一种用于在GPU上进行地震建模的方法包括:创建一个离散模型,该模型包含由具有任意边界条件的边界包围的体积,将离散化模型划分为子集,将数据传输到GPU,将存储在选定数量子集中的数据初始化 GPU执行与GPU上所选数量的子集的多个时间步长相关的计算,并且基本上同时计算从GPU卸载完成的子集数据,并将新的子集数据上载到GPU 。 提供的方法和相应的系统可以将存储在GPU存储器中的数据量减少一个数量级,从而最大限度地减少GPU设备到主机的传输带宽需求,并减少GPU设置时间。

    METHOD AND SYSTEM FOR CHECKPOINTING DURING SIMULATIONS
    3.
    发明申请
    METHOD AND SYSTEM FOR CHECKPOINTING DURING SIMULATIONS 有权
    在模拟期间检查的方法和系统

    公开(公告)号:US20110288831A1

    公开(公告)日:2011-11-24

    申请号:US13034341

    申请日:2011-02-24

    Abstract: Method and system for more efficient checkpointing strategy in cross correlating (316) a forward (328) and backward (308) propagated wave such as in migrating (326) or inverting seismic data. The checkpointing strategy includes storing in memory forward simulation data at a checkpointed time step, wherein the stored data are sufficient to do a cross correlation at that time step but not to restart the forward simulation. At other checkpoints, a greater amount of data sufficient to restart the simulation may be stored in memory (314). Methods are disclosed for finding an optimal combination, i.e. one that minimizes computation time (1132), of the two types of checkpoints for a given amount of computer memory (1004), and for locating a checkpoint at an optimal time step (306, 1214, 1310). The optimal checkpointing strategy (1002) also may optimize (1408) on use of fast (1402) vs. slow (1404) storage.

    Abstract translation: 交叉相关(316)前向(328)和向后(308)传播的波中的更有效的检查点策略的方法和系统,例如在迁移(326)或反转地震数据中。 检查点策略包括在检查点时间步骤中存储内存中的前向仿真数据,其中所存储的数据足以在该时间步骤进行互相关,但是不重新启动正向仿真。 在其他检查点,可以将足够重新启动模拟的更大量的数据存储在存储器(314)中。 公开了用于找到最佳组合的方法,即使给定量的计算机存储器(1004)的两种类型的检查点的计算时间(1132)最小化的最佳组合,并且用于在最佳时间步长(306,1214)处定位检查点 ,1310)。 最优检查点策略(1002)还可以优化(1408)快速(1402)与慢(1404)存储的使用。

    Method and System to Reduce: Memory Requirements, Device-to-Host Transfer Bandwidth Requirements, and Setup Time, for Seismic Modeling on Graphics Processing Units
    4.
    发明申请
    Method and System to Reduce: Memory Requirements, Device-to-Host Transfer Bandwidth Requirements, and Setup Time, for Seismic Modeling on Graphics Processing Units 有权
    减少图形处理单元的地震建模的方法和系统:内存要求,设备到主机传输带宽要求和建立时间

    公开(公告)号:US20120203523A1

    公开(公告)日:2012-08-09

    申请号:US13367189

    申请日:2012-02-06

    CPC classification number: G01V1/28 G06F9/5044 G06T1/20 G06T15/005 G09G5/363

    Abstract: A method for seismic modeling on a GPU includes creating a discretized model containing a volume enclosed by boundaries with arbitrary boundary conditions, dividing the discretized model into subsets, transferring a data to the GPU, initializing the data stored in a selected number of the subsets on the GPU, performing computation related with a number of time steps for the selected number of the subsets on the GPU, and substantially simultaneously computing, off-loading a finished subsets data from the GPU, and up-loading a new subsets data to the GPU. The provided method and corresponding system can reduce the amount of data stored in GPU memory by an order of magnitude, minimize the GPU device-to-host transfer bandwidth requirements, and reduce the GPU setup time.

    Abstract translation: 一种用于在GPU上进行地震建模的方法包括:创建一个离散模型,该模型包含由具有任意边界条件的边界包围的体积,将离散化模型划分为子集,将数据传输到GPU,将存储在选定数量子集中的数据初始化 GPU执行与GPU上所选数量的子集的多个时间步长相关的计算,并且基本上同时计算从GPU卸载完成的子集数据,并将新的子集数据上载到GPU 。 提供的方法和相应的系统可以将存储在GPU存储器中的数据量减少一个数量级,从而最大限度地减少GPU设备到主机的传输带宽需求,并减少GPU设置时间。

Patent Agency Ranking