Reducing I/O complexity by simulating coarse grained parallel algorithms

Dehne, Frank; Hutchinson, D. A. W.; Maheshwari, Anil; Dittrich, Wolfgang

doi:10.1109/ipps.1999.760428

Cited by 16 publications

(16 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are several papers on deriving external memory algorithms from parallel algorithms, by emulating the parallel machine on an external memory machine [25]- [28]. Most of these results also allow parallel disks [25], [27], [28] so that this is currently the main source of parallel disk external memory algorithms.…”

Section: Introductionmentioning

confidence: 99%

“…Most of these results also allow parallel disks [25], [27], [28] so that this is currently the main source of parallel disk external memory algorithms. Chiang et al [25] look at PRAM algorithms with special structure.…”

Section: Introductionmentioning

confidence: 99%

“…Chiang et al [25] look at PRAM algorithms with special structure. Dehne et al [27], [28] look at BSP algorithms and obtain efficient external algorithms for the special case of CGM algorithms. CGM algorithms work in supersteps and communicate their entire memory content in every superstep.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fast Concurrent Access to Parallel Disks

Sanders¹,

Egner²,

Korst³

2002

Algorithmica

111

View full text Add to dashboard Cite

High performance applications involving large data sets require the efficient and flexible use of multiple disks. In an external memory machine with D parallel, independent disks, only one block can be accessed on each disk in one I/O step. This restriction leads to a load balancing problem that is perhaps the main inhibitor for the efficient adaptation of single-disk external memory algorithms to multiple disks. We solve this problem for arbitrary access patterns by randomly mapping blocks of a logical address space to the disks.We show that a shared buffer of O(D) blocks suffices to support efficient writing. The analysis uses the properties of negative association to handle dependencies between the random variables involved. This approach might be of independent interest for probabilistic analysis in general.If two randomly allocated copies of each block exist, N arbitrary blocks can be read within N /D + 1 I/O steps with high probability. The redundancy can be further reduced from 2 to 1 + 1/r for any integer r without a big impact on reading efficiency. From the point of view of external memory models, these results rehabilitate Aggarwal and Vitter's "single-disk multi-head" model [1] that allows access to D arbitrary blocks in each I/O step. This powerful model can be emulated on the physically more realistic independent disk model [2] with small constant overhead factors. Parallel disk external memory algorithms can therefore be developed in the multi-head model first. The emulation result can then be applied directly or further refinements can be added.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fast Concurrent Access to Parallel Disks

Sanders¹,

Egner²,

Korst³

2002

Algorithmica

111

View full text Add to dashboard Cite

show abstract

“…In [10,12,17] schemes are presented that simulate parallel algorithms designed for coarsegrained parallel models, such as BSP [18], BSP* [4], CGM [11], on the EM model. The main intuition behind these works is that the interleaving between large local computation and bulk communication phases, which characterizes coarse-grained parallel algorithms, maps nicely on the twolevel structure of the EM model.…”

Section: Introductionmentioning

confidence: 99%

Translating submachine locality into locality of reference

Fantozzi¹,

Pietracaprina²,

Pucci³

2006

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

show abstract

“…Earlier results [8,10,15] provided evidence that efficient EM algorithms for two-level hierarchies can be obtained by simulating parallel ones written for coarse-grained parallel models, such as BSP [17] or CGM [9], on the EM model. The main intuition behind these works is that the interleaving be-tween large local computation and bulk communication phases, which characterizes coarse-grained parallel algorithms, maps nicely on the two-level structure of the EM model.…”

Section: Introductionmentioning

confidence: 99%

Cache-oblivious simulation of parallel programs

Pietracaprina¹,

Pucci²,

Silvestri³

2006

Proceedings 20th IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

This paper explores the relation between the structured parallelism exposed by the Decomposable BSP (D-BSP) model through submachine locality and locality of reference in multi-level cache hierarchies. Specifically, an efficient cache-oblivious algorithm is developed to simulate D-BSP programs on the Ideal Cache Model (ICM). The effectiveness of the simulation is proved by showing that optimal cache-oblivious algorithms for prominent problems can be obtained from D-BSP algorithms. Finally, a tight relation between optimality in the D-BSP and ICM models is established.

show abstract

Reducing I/O complexity by simulating coarse grained parallel algorithms

Abstract: Block-wise access to data is a central theme in the de

Cited by 16 publications

References 25 publications

Fast Concurrent Access to Parallel Disks

Fast Concurrent Access to Parallel Disks

Translating submachine locality into locality of reference

Cache-oblivious simulation of parallel programs

Contact Info

Product

Resources

About