The paper entitled “Pump Up the Volume: Processing Large Data on GPUs with Fast Interconnects,” by Clemens Lutz, Sebastian Breß, Steffen Zeuch, Tilmann Rabl (now at HPI), and Volker Markl explores the use of GPUs to accelerate database query processing.
GPUs are generally ill-suited for large-scale data processing for two reasons: (1) the on-board memory capacity is too small to store large datasets, and (2) the interconnect bandwidth to CPU main-memory is insufficient for ad-hoc data transfers. As a result, GPU-enabled systems face a data transfer bottleneck and do not scale to large datasets. In the paper, the authors investigate how a fast interconnect, such as NVLink 2.0 (linking dedicated GPUs to a CPU) can overcome the two scalability issues.
Consequently, in their experiments they demonstrate that a no-partitioning hash join achieves a speed-up of up to 18x over PCI-e 3.0 and up to 7.3x over an optimized CPU implementation. To download a preprint of the paper visit: https://bit.ly/3bHb3XO.