Revisiting co-processing for hash joins on the coupled CPU-GPU architecture

He, Jiong; Lu, Ming–Wei; He, Bingsheng

doi:10.14778/2536206.2536216

Cited by 107 publications

(77 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite the effectiveness of previous studies on query coprocessing on coupled architectures, both CPU and GPU execute homogeneous workloads in previous studies [19,7,38,21]. However, due to the unique architectural design of coupled CPU-GPU architectures, such homogeneous workload distribution schemes can hinder query co-processing performance on the GPU.…”

Section: Introductionmentioning

confidence: 95%

See 1 more Smart Citation

In-cache query co-processing on coupled CPU-GPU architectures

Zhang

2014

Proc. VLDB Endow.

Self Cite

View full text Add to dashboard Cite

Recently, there have been some emerging processor designs that the CPU and the GPU (Graphics Processing Unit) are integrated in a single chip and share Last Level Cache (LLC). However, the main memory bandwidth of such coupled CPU-GPU architectures can be much lower than that of a discrete GPU. As a result, current GPU query coprocessing paradigms can severely suffer from memory stalls. In this paper, we propose a novel in-cache query co-processing paradigm for main memory On-Line Analytical Processing (OLAP) databases on coupled CPU-GPU architectures. Specifically, we adapt CPU-assisted prefetching to minimize cache misses in GPU query co-processing and CPU-assisted decompression to improve query execution performance. Furthermore, we develop a cost model guided adaptation mechanism for distributing the workload of prefetching, decompression, and query execution between CPU and GPU. We implement a system prototype and evaluate it on two recent AMD APUs A8 and A10. The experimental results show that 1) in-cache query co-processing can effectively improve the performance of the state-of-the-art GPU co-processing paradigm by up to 30% and 33% on A8 and A10, respectively, and 2) our workload distribution adaption mechanism can significantly improve the query performance by up to 36% and 40% on A8 and A10, respectively.

show abstract

Section: Introductionmentioning

confidence: 95%

“…Coupled CPU-GPU architectures call for new data processing mechanisms. There have been studies on more collaborative and fine-grained schemes for query co-processing [19,38] and other data processing workloads (e.g., key-value stores [21] and MapReduce [7]). …”

Section: Introductionmentioning

confidence: 99%

In-cache query co-processing on coupled CPU-GPU architectures

Zhang

2014

Proc. VLDB Endow.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Since this survey should cover relational GDBMS, we only consider systems that are capable of using the GPU for most relational operations. That is, we disregard stand-alone approaches for accelerating a certain relational operator (e.g., He and others [30,32]), special co-processing techniques (e.g., Pirk and others [49]), or other -non data-processing related -applications for GPUs in database systems [33]. Furthermore, we will not discuss systems using other data models than the relational model, such as graph databases (e.g., Medusa from Zhong and He [64,65]) or MapReduce (e.g., Mars from He and others [28]).…”

Section: Research Questionsmentioning

confidence: 99%

“…Access structures and algorithms for the GPU: In order to support GPUacceleration, a DBMS needs to implement access structures on the GPU (e.g., columns or B + -trees) and operators that work on them. Here, the most approaches were developed [7,21,29,32,48,49]. Data Placement Strategy: A DBMS needs to keep track of which data is stored on the GPU, and which access structure needs to be transferred to GPU memory [29].…”

Section: Summary: Extension Points For Main-memory Dbmssmentioning

confidence: 99%

“…This development inspired the database research community to investigate methods for accelerating database systems via GPU co-processing. Several research papers and performance studies demonstrate the potential of this approach [7,21,29,32,48,49] -and the technology has also found its way into commercial products (e.g., Jedox [1] or ParStream [2]). …”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GPU-Accelerated Database Systems: Survey and Open Challenges

Breß

Heimel

Siegmund

et al. 2014

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The vast amount of processing power and memory bandwidth provided by modern graphics cards make them an interesting platform for data-intensive applications. Unsurprisingly, the database research community identified GPUs as effective co-processors for data processing several years ago. In the past years, there were many approaches to make use of GPUs at different levels of a database system. In this paper, we explore the design space of GPU-accelerated database management systems. Based on this survey, we present key properties, important trade-offs and typical challenges of GPU-aware database architectures, and identify major open challenges. Additionally, we survey existing GPU-accelerated DBMSs and classify their architectural properties. Then, we summarize typical optimizations implemented in GPU-accelerated DBMSs. Finally, we propose a reference architecture, indicating how GPU acceleration can be integrated in existing DBMSs.

show abstract

An evaluation of analytical queries on CPUs and coupled GPUs

Luan

Chang

2016

Concurrency and Computation

View full text Add to dashboard Cite

Summary Recently, the mainstream hardware vendors such as Intel and AMD have made significant efforts to integrate the central processing unit (CPU) and the graphics processing unit (GPU) into a single chip, which forms a coupled CPU‐GPU architecture. Data transfer between the CPU and the GPU through a Peripheral Component Interconnect Express bus is eliminated on this architecture, which provides new opportunities for database community to optimize query processing. Because of the lack of comprehensive evaluation of database systems on coupled CPU‐GPU platforms, it is difficult for academic and industry researchers to make appropriate decisions on improvement and optimization directions. In this paper, we conduct an extensive experimental study to evaluate an online analytical processing system on Intel and AMD machines. The performance difference is measured and analyzed when executing queries on integrated GPUs and multicore CPUs. The impacts of various parameters, data sizes, and optimization techniques on performance are also investigated. The results provide preliminary insights into database query and operator behaviors on state‐of‐the‐art coupled CPU‐GPU architectures.

show abstract

Revisiting co-processing for hash joins on the coupled CPU-GPU architecture

Cited by 107 publications

References 24 publications

In-cache query co-processing on coupled CPU-GPU architectures

In-cache query co-processing on coupled CPU-GPU architectures

GPU-Accelerated Database Systems: Survey and Open Challenges

An evaluation of analytical queries on CPUs and coupled GPUs

Contact Info

Product

Resources

About