Parallel Search On Video Cards

Abstract:  Recent approaches exploiting the massively parallel architecture of graphics processors (GPUs) to accelerate database operations have achieved intriguing results. We present a novel parallel search algorithm for large-scale database index operations that outperforms traditional thread-level parallel GPU and CPU implementations. With parallel architectures becoming omnipresent, and with searching being a fundamental functionality for many applications, we expect it to be applicable beyond the database domain. While the GPUs do not appear to be ready to be adopted for general-purpose database applications yet, mostly due to their batch processing mode that implies slow response times, given their rapid development, this will change in the near future. The trend towards combining CPU and GPU processing encourages development of parallel techniques on either architecture.

Tim Kaldewey is a Researcher at Oracle Corporation and a 4-year PhD

student at UCSC's Systems Research Group. He previously held positions

at IBM, SAP, Lufthansa and SAG. His research focuses on high performance

data management, in particular, parallel algorithms for emerging architectures and predictable resource management.