Efficient Unsteady Flow Visualization with High-Order Access Dependencies


We present a novel high-order access dependencies-based model for efficient pathline computation in unsteady flow visualization. By taking longer access sequences into account to model more sophisticated data access patterns in particle tracing, our method greatly improves the accuracy and reliability in data access prediction. In our work, high-order access dependencies are calculated by tracing uniformly seeded pathlines in both forward and backward directions in a preprocessing stage. The effectiveness of our approach is demonstrated through a parallel particle tracing framework with high-order data prefetching. Results show that our method achieves higher data locality and hence improves the efficiency of pathline computation.
Keywords: Flow visualization, Pathline computation, High-order access dependencies


Figure 1. The pipeline of our work. In the preprocessing stage, particles are uniformly seeded and traced by the input of the raw flow data. The highorder access dependencies are further computed according to the generated pathlines. We further integrate high-order access dependencies into data blocks. A parallel particle tracing framework that performs high-order data prefetching is employed to demonstrate that our method achieves better efficiency than that of the first-order method.

Figure 2. The computation of high-order access dependencies. From left to right: The nine pathlines originated from block (2, 1); The graph model of access dependencies recorded in these pathlines; The 3rd-order “accumulated” access dependencies of block (2, 1), which include all access dependencies with 1st, 2nd, and 3rd order. At the bottom, the access transition probability is defined as the ratio between the number of pathlines traveling to this next block and the total number of pathlines traveling from the same historical blocks to all possible next blocks.


Jiang Zhang, Hanqi Guo, and Xiaoru Yuan. Efficient Unsteady Flow Visualization with High-Order Access Dependencies. In Proceedings of IEEE Pacific Visualization Symposium (PacificVis 2016), pages 80-87, Taipei, Apr. 19-22, 2016.


Title     = {Efficient Unsteady Flow Visualization with High-Order Access Dependencies},
Author    = {Jiang Zhang and Hanqi Guo and Xiaoru Yuan},
Booktitle = {Proceedings of {IEEE} Pacific Visualization Symposium 2016},
Year      = {2016},
Pages     = {80--87},
Bibsource = {dblp computer science bibliography, http://dblp.org}