-
公开(公告)号：US20130232476A1
公开(公告)日：2013-09-05
申请号：US13409536
申请日：2012-03-01
IPC分类号： G06F9/45
摘要： A system and associated method for automatically pipeline parallelizing a nested loop in sequential code over a predefined number of threads. Pursuant to task dependencies of the nested loop, each subloop of the nested loop are allocated to a respective thread. Combinations of stage partitions executing the nested loop are configured for parallel execution of a subloop where permitted. For each combination of stage partitions, a respective bottleneck is calculated and a combination with a minimum bottleneck is selected for parallelization.
摘要翻译： 一种系统和相关联的方法，用于在预定义数量的线程上自动流水线化序列码中的嵌套循环。 根据嵌套循环的任务依赖关系，将嵌套循环的每个子循环分配给相应的线程。 执行嵌套循环的分段分组的组合被配置为在允许的情况下并行执行子循环。 对于级分段的每个组合，计算相应的瓶颈，并且选择具有最小瓶颈的组合用于并行化。
-
公开(公告)号：US08949809B2
公开(公告)日：2015-02-03
申请号：US13409536
申请日：2012-03-01
IPC分类号： G06F9/45
摘要： A system and associated method for automatically pipeline parallelizing a nested loop in sequential code over a predefined number of threads. Pursuant to task dependencies of the nested loop, each subloop of the nested loop are allocated to a respective thread. Combinations of stage partitions executing the nested loop are configured for parallel execution of a subloop where permitted. For each combination of stage partitions, a respective bottleneck is calculated and a combination with a minimum bottleneck is selected for parallelization.
摘要翻译： 一种系统和相关联的方法，用于在预定义数量的线程上自动流水线化序列码中的嵌套循环。 根据嵌套循环的任务依赖关系，将嵌套循环的每个子循环分配给相应的线程。 执行嵌套循环的分段分组的组合被配置为在允许的情况下并行执行子循环。 对于级分段的每个组合，计算相应的瓶颈，并且选择具有最小瓶颈的组合用于并行化。
-
公开(公告)号：US08910137B2
公开(公告)日：2014-12-09
申请号：US13446089
申请日：2012-04-13
IPC分类号： G06F9/44
CPC分类号： G06F8/45 , G06F8/30 , G06F8/4452 , G06F8/452 , G06F8/72 , G06F11/3466 , G06F2201/865
摘要： A system and associated method for creating a library method stub in source code form corresponding to an original library call in machine-executable form. Sequential code subject to pipeline-parallelization includes the original library call of which source code is not available for profiling. The original library call is annotated for a library method signature to specify formal arguments and parameters accessible to the formal arguments. Memory operations of the original library call are summarized with on one or more idiom sentences, which are mapped to a respective code stub that simulates the original library call. The library method stub is created for profiling of the original library call in source code level such that the sequential code is parallelized via the created library method stub.
摘要翻译： 用于以与机器可执行形式的原始库调用相对应的源代码形式创建库方法存根的系统和关联方法。 流水线并行化的顺序代码包括原始库调用，哪个源代码不可用于分析。 原始库调用被注释为库方法签名，以指定形式参数可访问的形式参数和参数。 原始库调用的存储器操作在一个或多个成语句上被概括，其被映射到模拟原始库调用的相应代码存根。 创建库方法存根，用于在源代码级别对原始库调用进行概要分析，以便顺序代码通过创建的库方法存根进行并行化。
-
公开(公告)号：US20130275955A1
公开(公告)日：2013-10-17
申请号：US13446089
申请日：2012-04-13
IPC分类号： G06F9/44
CPC分类号： G06F8/45 , G06F8/30 , G06F8/4452 , G06F8/452 , G06F8/72 , G06F11/3466 , G06F2201/865
摘要： A system and associated method for creating a library method stub in source code form corresponding to an original library call in machine-executable form. Sequential code subject to pipeline-parallelization includes the original library call of which source code is not available for profiling. The original library call is annotated for a library method signature to specify formal arguments and parameters accessible to the formal arguments. Memory operations of the original library call are summarized with on one or more idiom sentences, which are mapped to a respective code stub that simulates the original library call. The library method stub is created for profiling of the original library call in source code level such that the sequential code is parallelized via the created library method stub.
摘要翻译： 用于以与机器可执行形式的原始库调用相对应的源代码形式创建库方法存根的系统和关联方法。 流水线并行化的顺序代码包括原始库调用，哪个源代码不可用于分析。 原始库调用被注释为库方法签名，以指定形式参数可访问的形式参数和参数。 原始库调用的存储器操作在一个或多个成语句上被概括，其被映射到模拟原始库调用的相应代码存根。 创建库方法存根，用于在源代码级别对原始库调用进行概要分析，以便顺序代码通过创建的库方法存根进行并行化。
-
-
-
搜索结果
4 条记录 (耗时 0.037 秒)
