Exploiting Massive Parallelism for IndexingMulti-Dimensional Datasets on the GPU

Kim, Jinwoong; Jeong, Won-Ki; Nam, Beomseok

doi:10.1109/tpds.2014.2347041

Cited by 26 publications

(16 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Indexing On the Gpumentioning

confidence: 99%

“…Kim et al [28] designed an R-tree for the GPU to optimize index searches that avoids many of the drawbacks of executing tree traversals on the GPU. Later, the same research group presented a hybrid approach [29] that splits the R-tree between the CPU and GPU by assigning parts of the algorithm with more regular and irregular instruction ows to the GPU and CPU, respectively.…”

Section: Indexing On the Gpumentioning

confidence: 99%

“…To reduce the number of comparisons between data points when ltering the candidate set, indexes are used. However, in high dimensionality, index searches become increasingly exhaustive due to the well-known curse of dimensionality [7,16,28,37], where the index needs to search a large fraction of the database, potentially degrading the search to the performance level obtained by a brute force search. us, searches in high dimensionality can be prohibitive.…”

Section: Introductionmentioning

confidence: 99%

“…ere are two major indexing strategies for the GPU: (i) index-trees, similar to those that have been shown to provide good performance on the CPU, such as the R-tree [24]; or (ii) non-hierarchical indexes, such as grids or binning. Several works propose e cient indexes for points or other objects on the GPU [12,20,21,28,29,38].Kim et al [28] designed an R-tree for the GPU to optimize index searches that avoids many of the drawbacks of executing tree traversals on the GPU. Later, the same research group presented a hybrid approach [29] that splits the R-tree between the CPU and GPU by assigning parts of the algorithm with more regular and irregular instruction ows to the GPU and CPU, respectively.…”

mentioning

confidence: 99%

See 3 more Smart Citations

GPU-Accelerated Similarity Self-Join for Multi-Dimensional Data

Gowanlock

Karsin

2019

Proceedings of the 15th International Workshop on Data Management on New Hardware

View full text Add to dashboard Cite

e self-join nds all objects in a dataset that are within a search distance, ϵ , of each other; therefore, the self-join is a building block of many algorithms. We advance a GPU-accelerated self-join algorithm targeted towards high dimensional data. e massive parallelism a orded by the GPU and high aggregate memory bandwidth makes the architecture well-suited for data-intensive workloads. We leverage a grid-based, GPU-tailored index to perform range queries. We propose the following optimizations: (i) a trade-o between candidate set ltering and index search overhead by exploiting properties of the index; (ii) reordering the data based on variance in each dimension to improve the ltering power of the index; and (iii) a pruning method for reducing the number of expensive distance calculations. Across most scenarios on real-world and synthetic datasets, our algorithm outperforms the parallel state-of-the-art approach. Exascale systems are converging on heterogeneous distributed-memory architectures. We show that an entity partitioning method can be utilized to achieve a balanced workload, and thus good scalability for multi-GPU or distributed-memory self-joins.us, many large-scale data analytics applications will rely on GPU-e cient algorithms, including the distance similarity self-join for high dimensional data -the subject of this work. is paper makes the following novel contributions:• Leveraging an e cient indexing scheme for the GPU, we exploit the trade-o between index ltering power and search cost to improve the overall performance of searching high dimensional feature spaces.• We improve the ltering power of the index by reordering the data in each dimension using statistical properties of the data distribution. We show that this is particularly important when exploiting the trade-o outlined above.• We mitigate the performance cost of reducing index ltering power by proposing a technique that prunes the candidate set by comparing points based on an un-indexed dimension.• We show that on the worst-case data distribution for our approach, we achieve signi cantly be er performance than the state-of-the-art on the same scenario. is suggests that the performance of the GPU-accelerated self-join is resilient to the data distribution, making the approach well-suited for many application scenarios.• We evaluate our approach on 5 real-world and 3 synthetic datasets and show that our GPU accelerated self-join outperforms the state-of-the-art parallel algorithm in the literature.• e self-join is an expensive operation. We show initial insights into the scalability of the self-join on multi-GPU and distributed-memory systems, and demonstrate that an entity partitioning strategy can be used to achieve good load balancing. e paper is outlined as follows: Section 2 provides background material, Section 3 formalizes the problem and discusses previous work that we employ, Section 4 presents the novel methods we use to improve high dimensional self-join performance, Section 5 illustrates our performance results, Section 6 dis...

show abstract

Section: Indexing On the Gpumentioning

confidence: 99%

Section: Indexing On the Gpumentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

GPU-Accelerated Similarity Self-Join for Multi-Dimensional Data

Gowanlock

Karsin

2019

Proceedings of the 15th International Workshop on Data Management on New Hardware

View full text Add to dashboard Cite

show abstract

“…The disadvantage of index-trees is that they contain many branch instructions, which can reduce the parallel efficiency of the GPU due to the SIMD architecture. An R-tree for the GPU was presented in [23], which was optimized to reduce thread divergence. Later, the same authors showed that it is better to perform the tree traversal on the CPU and perform the scanning of the leaf nodes on the GPU [10].…”

Section: Introductionmentioning

confidence: 99%