CUDA Based Parallel Implementations of Space-Saving on a GPU

被引:14
作者
Cafaro, Massimo [1 ,2 ]
Epicoco, Italo [1 ,2 ]
Aloisio, Giovanni [1 ,2 ]
Pulimeno, Marco [3 ]
机构
[1] Univ Salento, Dept Engn Innovat, Lecce, Italy
[2] Euromediterranean Ctr Climate Change Fdn, Rome, Italy
[3] Univ Salento, Dept Math & Phys, Lecce, Italy
来源
2017 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS) | 2017年
关键词
FREQUENT;
D O I
10.1109/HPCS.2017.108
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
We present four CUDA based parallel implementations of the Space-Saving algorithm for determining frequent items on a GPU. The first variant exploits the open-source CUB library to simplify the implementation of a user's defined reduction, whilst the second is based on our own implementation of the parallel reduction. The third and the fourth, built on the previous variants, are meant to improve the performance by taking advantage of hardware based atomic instructions. In particular, we implement a warp based ballot mechanism to accelerate the Space-Saving updates. We show that our implementation of the parallel reduction, coupled with the ballot based update mechanism, is the fastest, and provides extensive experimental results regarding its performance.
引用
收藏
页码:707 / 714
页数:8
相关论文
共 17 条
[1]  
[Anonymous], 2009, PVLDB
[2]  
[Anonymous], 2017, CUDA toolkit documentation
[3]  
Cafaro M., CONCURRENCY IN PRESS
[4]   A parallel space saving algorithm for frequent items and the Hurwitz zeta distribution [J].
Cafaro, Massimo ;
Pulimeno, Marco ;
Tempesta, Piergiulio .
INFORMATION SCIENCES, 2016, 329 :1-19
[5]   Finding frequent items in parallel [J].
Cafaro, Massimo ;
Tempesta, Piergiulio .
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2011, 23 (15) :1774-1788
[6]  
Charikar M, 2002, LECT NOTES COMPUT SC, V2380, P693
[7]   An improved data stream summary: the count-min sketch and its applications [J].
Cormode, G ;
Muthukrishnan, S .
JOURNAL OF ALGORITHMS-COGNITION INFORMATICS AND LOGIC, 2005, 55 (01) :58-75
[8]   Methods for finding frequent items in data streams [J].
Cormode, Graham ;
Hadjieleftheriou, Marios .
VLDB JOURNAL, 2010, 19 (01) :3-20
[9]  
Demaine ED, 2002, LECT NOTES COMPUT SC, V2461, P348
[10]   Frequent Items Mining Acceleration Exploiting Fast Parallel Sorting on the GPU [J].
Erra, Ugo ;
Frola, Bernardino .
PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE, ICCS 2012, 2012, 9 :86-95