Optimization of Parallel Streamline Visualization Algorithm Based on Vtk-m
Streamlines are one of the most illuminating techniques to achieve fundamental goal of scientific insight from result-ing numerical simulations.Due to the large scale of the flow field data,the traditional streamline algorithm is inefficient,resulting in a slow visualization process.Although the performance of the streamline algorithm provided in the open source vtk-m has been im-proved,the parallel efficiency is low under multithreading.To solve this problem,a parallel streamline visualization optimization al-gorithm based on dynamic node tree is proposed.Mainly through two divisions of coarse-grained and fine-grained,parallel construc-tion of dynamic node tree to organize and manage data,and the index structure is established to conveniently narrow the alternative areas in the integration process.To reduce memory requirements of streamline integration by the methods of data blocks attribute ex-traction and employ parallel execution of seed integral tasks to leverage multi-core computing resources.The experimental results under different scale data sets show the effectiveness of the optimized algorithm.
万方数据
维普
