GPGPU-accelerated interesting interval discovery and other computations on GeoSpatial datasets

Abstract

It is imperative that for scalable solutions of GIS computations the modern hybrid architecture comprising a CPU-GPU pair is exploited fully. The existing parallel algorithms and data structures port reasonably well to multicore CPUs, but poorly to GPGPUs because of latter's atypical fine-grained, single-instruction multiple-thread (SIMT) architecture, extreme memory hierarchy and coalesced access requirements, and delicate CPU-GPU coordination. Recently, our parallelization of the state-of-art interesting sequence discovery algorithms calculates one-dimensional interesting intervals over an image representing the normalized di↵erence vegetation indices of Africa within 31 ms on an nVidia 480GTX. To our knowledge, this paper reports the first parallelization of these algorithms. This allowed us to process 612 images representing biweekly data from July 1981 through Dec 2006 within 22 seconds. We were also able to pipe the output to a display in almost real-time, which would interest climate scientists. We have also undertaken parallelization of two key tree-based data structures, namely R-tree and heap, and have employed parallel R-tree in polygon overlay system. These data structure parallelization are hard because of the underlying tree topology and the finegrained computation leading to frequent access to such data structures severely stifling parallel e ciency.

Interesting Spatio-Temporal Region Discovery Computations Over Gpu and Mapreduce Platforms

“…This quickly becomes problematic with today's large data sets. (Zhou et al, 2011) provides an SEP algorithm which reduces computation by an order of magnitude over naive implementations through the construction of a lookup (Zhou et al, 2011, Prasad et al, 2013a. Each column for the purposes of the discovery is assumed independent.…”

“…In our previous work on interesting path discovery, the GPU implementation is straightforward due to the data independent nature of the problem (Prasad et al, 2013a). The lookup table is embarrassingly parallel to implement by launching a thread for every n ∈ N .…”

“…A worse case O(N 2 M ) problem is reduced in theoretical complexity to O(N M ) through parallelization on the GPU (Prasad et al, 2013a).…”

“…Explicit elimination programatically removes smaller subpaths that are completely embedded in larger subpaths. The complexity of explicit elimination is the same as with discovery (Prasad et al, 2013a).…”

“…We were able to process the entire data set quickly enough to generate real-time visualizations of the algorithm's output on a single GPU (Prasad et al, 2013b, Prasad et al, 2013a. We use this data set and GPU implementation in Section 4.1.…”