Clustering billions of data points using GPUs

Ren, Wu; Zhang, Bin; Hsu, Meichun

doi:10.1145/1531666.1531668

Cited by 81 publications

(38 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As pointed out in [17], bitmap approach is elegant in expressing the problem, but it is not a good method for high performance, since bitmap takes more space when k is large and requires more shared memory.…”

Section: Gpuminermentioning

confidence: 99%

“…HP_k-Means is by far the most efficient k-Means algorithm on GPUs [17]. However, the details of HP_k-Means are not reported in the paper.…”

Section: Hp_kmeansmentioning

confidence: 99%

“…The iteration time is fifty, and the algorithm will stop after fifty rounds, no matter what the change of the centroid is. Besides, the speeds of HP_k-Means, UV_k-Means and GPUMiner are extracted from [17]. As we use the same experimental configurations as [17], our comparisons are fair and reasonable.…”

Section: A On Low-dimensional Data Setsmentioning

confidence: 99%

“…Some clustering algorithms have also been implemented on the GPUs, including k-Means. There are mainly three existing GPU-based k-Means algorithms: GPUMiner [15], UV_k-Means [16], and HP_k-Means [17]. UV_k-Means achieves a speedup of ten to forty as compared with a four-threaded Minebench [18] running on a dual-core, hyper-threaded CPU.…”

Section: Introductionmentioning

confidence: 99%

“…UV_k-Means achieves a speedup of ten to forty as compared with a four-threaded Minebench [18] running on a dual-core, hyper-threaded CPU. HP_k-Means claims another speedup of two to four as compared with UV_k-Means and twenty to seventy speedup as compared with GPUMiner [17]. These existing works have shown the promising high performance advantage of GPUs.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Speeding up K-Means Algorithm by GPUs

Zhao

Chu

et al. 2010

2010 10th IEEE International Conference on Computer and Information Technology

View full text Add to dashboard Cite

Abstract-Cluster analysis plays a critical role in a wide variety of applications, but it is now facing the computational challenge due to the continuously increasing data volume. Parallel computing is one of the most promising solutions to overcoming the computational challenge. In this paper, we target at parallelizing k-Means, which is one of the most popular clustering algorithms, by using the widely available Graphics Processing Units (GPUs). Different from existing GPU-based k-Means algorithms, we observe that data dimension is an important factor that should be taken into consideration when parallelizing k-Means on GPUs. In particular, we use two different strategies for low-dimensional data sets and high-dimensional data sets respectively, in order to make the best use of the power of GPUs. For lowdimensional data sets, we exploit GPU on-chip registers to significantly decrease data access latency. For highdimensional data sets, we design a novel algorithm which simulates matrix multiplication and exploits GPU on-chip registers and also on-chip shared memory to achieve high compute-to-memory-access ratio. As a result, our GPU-based k-Means algorithm is three to eight times faster than the best reported GPU-based algorithm.

show abstract

Section: Gpuminermentioning

confidence: 99%

“…HP_k-Means is by far the most efficient k-Means algorithm on GPUs [17]. However, the details of HP_k-Means are not reported in the paper.…”

Section: Hp_kmeansmentioning

confidence: 99%

Section: A On Low-dimensional Data Setsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Speeding up K-Means Algorithm by GPUs

Zhao

Chu

et al. 2010

2010 10th IEEE International Conference on Computer and Information Technology

View full text Add to dashboard Cite

show abstract

Finding exemplars in dense data with affinity propagation on clusters of GPUs

Kurdziel

Boryczko

2012

Concurrency and Computation

View full text Add to dashboard Cite

This work presents an efficient implementation of affinity propagation (AP) on clusters of graphical processing units (GPUs). AP is a state-of-the-art method for finding exemplars in data sets described by similarity matrices. It is typically employed in crisp clustering applications. However, when finding exemplars in an n-pattern data set with dense, non-metric similarities, AP performs iterative processing of three n n floating point matrices. One of them stores the similarities, and the other two store the values that will ultimately pinpoint the exemplars. For large similarity matrices, AP is therefore computationally expensive. Although matrix operations of AP are well suited for GPUs, its memory footprint limits the size of tasks that can be solved on one unit. We present, however, a decomposition scheme for AP that distributes the calculations over multiple GPUs, with low communication-to-computation ratio. Because of this favorable communication pattern, our implementation finds exemplars in large, dense similarity data efficiently, even when GPUs are connected by a slow network. Furthermore, by combining global device memory of multiple GPUs, it can solve problems that would not fit in a single unit.

show abstract

A fast parallel clustering algorithm for molecular simulation trajectories

et al. 2012

View full text Add to dashboard Cite

We implemented a GPU-powered parallel k-centers algorithm to perform clustering on the conformations of molecular dynamics (MD) simulations. The algorithm is up to two orders of magnitude faster than the CPU implementation. We tested our algorithm on four protein MD simulation datasets ranging from the small Alanine Dipeptide to a 370-residue Maltose Binding Protein (MBP). It is capable of grouping 250,000 conformations of the MBP into 4000 clusters within 40 seconds. To achieve this, we effectively parallelized the code on the GPU and utilize the triangle inequality of metric spaces. Furthermore, the algorithm's running time is linear with respect to the number of cluster centers. In addition, we found the triangle inequality to be less effective in higher dimensions and provide a mathematical rationale. Finally, using Alanine Dipeptide as an example, we show a strong correlation between cluster populations resulting from the k-centers algorithm and the underlying density. © 2012 Wiley Periodicals, Inc.

show abstract

Clustering billions of data points using GPUs

Cited by 81 publications

References 1 publication

Speeding up K-Means Algorithm by GPUs

Speeding up K-Means Algorithm by GPUs

Finding exemplars in dense data with affinity propagation on clusters of GPUs

A fast parallel clustering algorithm for molecular simulation trajectories

Contact Info

Product

Resources

About