并行程序设计Chapter3资料.ppt

第3章 易并行计算Embarrassingly Parallel Computations 1、IDEAL PARALLEL COMPUTATION 2、EMBARRASSINGLY PARALLEL EXAMPLES Geometrical Transformation of Images 图形的几何变 换 Mandelbrot Set 曼德勃罗特集 Monte Carlo Methods 蒙特卡罗法 3.1 IDEAL PARALLEL COMPUTATION Embarrassingly Parallel Computations A computation that can obviously be divided into a number of completely independent parts, each of which can be executed by a separate process(or). This means there are no communications between every process. 3.1 IDEAL PARALLEL COMPUTATION No communication or very little communication between processes Each process can do its tasks without any interaction with other processes 3.1 IDEAL PARALLEL COMPUTATION Nearly embarrassingly parallel computations are those that require results to be distributed and combined in some way. This means only one process is running at first and at last. Master-slave structure Static process creation Dynamic process creation 3.1 IDEAL PARALLEL COMPUTATION Practical embarrassingly parallel computation with static process creation and master-slave approach 3.1 IDEAL PARALLEL COMPUTATION Practical embarrassingly parallel computation with dynamic process creation and master-slave approach 3.2 EMBARRASSINGLY PARALLEL EXAMPLES 1、Geometrical Transformation of Images 图形的几何变 换 2、Mandelbrot Set 曼德勃罗特集 3、Monte Carlo Methods 蒙特卡罗法 3.2.1 Geometrical Transformation of Images Many low level image processing operations only involve local data with very limited if any communication between areas of interest. 3.2.1 Geometrical Transformation of Images 3.2.1 Geometrical Transformation of Images 3.2.1 Geometrical Transformation of Images 3.2.1 Geometrical Transformation of Images 并行程序 —— 易并行计算 主从结构，一个主进程，48个从进程。 主进程向每个从进程发送需处理的10 行的首行号，并接收每个从进程的处理结 果以供显示。 每个从进程处理10*640的区域。 3.2.1 Geometrical Transformation