Wang, ChaoliGao, JinzhuLi, LiyaShen, Han-WeiKlaus Mueller and Thomas Ertl and Eduard Groeller2014-01-292014-01-2920053-905673-26-61727-8376https://doi.org/10.2312/VG/VG05/011-019We present a new parallel multiresolution volume rendering framework for large-scale time-varying data visualization using the wavelet-based time-space partitioning (WTSP) tree. Utilizing the wavelet transform, a largescale time-varying data set is converted into a space-time multiresolution data hierarchy, and is stored in a timespace partitioning (TSP) tree. To eliminate the parent-child data dependency for reconstruction and achieve loadbalanced rendering, we design an algorithm to partition the WTSP tree and distribute the wavelet-compressed data along hierarchical space-filling curves with error-guided bucketization. At run time, the WTSP tree is traversed according to the user-specified time step and tolerances of both spatial and temporal errors. Data blocks of different spatio-temporal resolutions are reconstructed and rendered to compose the final image in parallel. We demonstrate that our algorithm can reduce the run-time communication cost to a minimum and ensure a well-balanced workload among processors when visualizing gigabytes of time-varying data on a PC cluster.Categories and Subject Descriptors (according to ACM CCS): E.4 [Coding and Information Theory]: Data compaction and compression; I.3.1 [Computer Graphics]: Parallel processing; I.3.3 [Computer Graphics]: Picture and Image GenerationViewing algorithms; I.3.6 [Computer Graphics]: Methodology and TechniquesGraphics data structures and data types