Distributed Join Algorithms on Multi-CPU Clusters with GPUDirect RDMA

Guo, Chengxin; Chen, Hong; Zhang, Feng; Li, Cuiping

doi:10.1145/3337821.3337862

Cited by 4 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MG-Join: A Scalable Join for Massively Parallel Multi-GPU Architectures We further conduct in-depth experiments to demonstrate the efficiency of MG-Join in joining large data sets. Our experiments find that 1) MG-Join achieves up to 97% utilization of the bisection bandwidth of the GPU interconnect links, and 2) MG-Join achieves up to 2.5x performance improvement over existing multi-GPU hash join implementations[32,212]. Moreover, MG-Join helps improve the overall performance of TPC-H queries by up to 4.5x over multi-GPU version of an open-source commercial GPU database, Omnisci[7].The rest of this chapter is organized as follows.…”

mentioning

confidence: 81%

“…• In Section 5.4.3, we compare the overall performance of MG-Join against UMJ [32] and DPRJ [212]. For communication across GPUs within the same node, DPRJ makes use of CUDA communication APIs that chooses the direct routes between GPUs for data transfer.…”

Section: Methodsmentioning

confidence: 99%

“…Lutz et al [44] studied the performance of simple hash join operation and well as selection-aggregation queries on such hardware. Guo et al [212] proposed a hash join implementation for distributed RDMA clusters with…”

Section: Multi-gpu Architecturementioning

confidence: 99%

“…To demonstrate this, we study the performance of hash join, whose input relations are equally distributed to multiple GPUs. Specifically, we study two hash join designs for multi-GPU architectures, including UMJ [32] and DPRJ [212]. UMJ makes use of the NVIDIA unified memory feature to move the data between the GPUs, and DPRJ is designed based on the current state-of-the-art join implementation available for distributed GPU architectures.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

In-memory analytical query processing on GPUs

Paul¹

View full text Add to dashboard Cite

The high global memory bandwidth and the large number of parallel cores available in modern Graphics Processing Units (GPUs) make them ideal for highperformance Online Analytical Processing (OLAP) systems. However, it is challenging to design efficient high-performance query processing systems for GPUs due to the following reasons: 1) the rapid evolution of GPU hardware in the last decade, 2) the significant differences in the hardware architecture of GPUs when compared to CPUs, 3) the high overhead of moving the data between the CPU and GPU and 4) the small global memory size of a single GPU that necessitates the access of remote data over PCIe or NVLink interconnects when processing large data sets. In this thesis, we study existing query processing systems for GPUs and propose techniques to improve query execution efficiency on GPUs.We begin by studying the performance of hash join, which is one of the most compute and memory intensive relational operator in OLAP systems. Specifically, we first revisit the major GPU hash join implementations in the past decade that were designed to execute on a single GPU. We then detail how these implementations take advantage of different GPU architecture features and conduct a comprehensive evaluation of their performance and cost-efficiency using different generations of GPUs. This helps shed light on the impact of different GPU architecture features on the performance of the hash join operation and identify the factors guiding the choice of these features. We further study how data characteristics like skew and match rate impact the performance of GPU hash join implementations. Novel techniques to improve the performance of the hash join operation when joining input relations with severe skew or high match rate were also proposed as part of this study. Our evaluation finds that the proposed techniques help avoid any performance degradation when joining skewed input relations and achieve up to 2.5x better performance when joining input relations with high match rate.Next, we extend our study on the hash join operation to modern multi-GPU architectures. The recent scale-up of GPU hardware through the integration of multiple xi Contentsxii GPUs into a single machine and the introduction of higher bandwidth interconnects like NVLink 2.0 has enabled new opportunities of relational query processing on multiple GPUs. However, due to the unique characteristics of GPUs and the interconnects, existing hash join implementations spend up to 66% of their execution time moving the data between the GPUs and achieve lower than 50% utilization of the newer high bandwidth interconnects. This leads to extremely poor scalablity of hash join performance on multiple GPUs, which can be slower than the performance on a single GPU. In this thesis, we propose MG-Join, a scalable partitioned hash join implementation on multiple GPUs of a single machine. In order to effectively improve the bandwidth utilization, we develop a novel multi-hop routing for cross-GPU communication that adaptively chooses...

show abstract

mentioning

confidence: 81%

Section: Methodsmentioning

confidence: 99%

Section: Multi-gpu Architecturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations