Database Systems and Information Management

News

All news

DIMA Research Paper Accepted for Publication at EDBT 2024

We are pleased to announce that the research paper “Benchmarking Stream Join Algorithms on GPUs: A Framework and its Application to the State of the Art,” authored by Dwi Prasetyo Adi Nugroho, Philipp M. Grulich, Steffen Zeuch, Clemens Lutz, Stefano Bortoli, and Volker Markl has been accepted for publication at EDBT 2024 [1], the 27th InternaJonal Conference on Extending Database Technology to be held on March 25-28, 2024 in Paestum, Italy.

Title:
Benchmarking Stream Join Algorithms on GPUs: A Framework and its Application to the State of the Art

Authors:
Dwi P. A. Nugroho, Philipp M. Grulich, Steffen Zeuch, Clemens Lutz, Stefano Bortoli, Volker Markl
 

Abstract: 
Joining streams is an essential and resource-demanding operation in stream processing engines. Recent works have shown significant performance benefits by offloading stream-join processing to hardware accelerators, including GPUs. As a result, a wide variety of GPU-accelerated stream join algorithms (SJAs) have emerged. However, existing works evaluate the proposed GPU-accelerated SJAs only in isolation, on different hardware, and not using a common workload. Consequently, it is difficult to compare different SJAs and select the best-suited SJA for a particular situation. In our paper, we shed light on the performance characterisJcs of GPU-accelerated SJAs. Our experiments show that each variant of SJA has its strengths and weaknesses and that ill-suited configurations of parameters lead to up to two orders of magnitude difference in throughput. Based on our results, we have developed a guideline for selecting SJA variants for different circumstances.


[1] EDBT 2024, https://dastlab.github.io/edbticdt2024/