TY - GEN
T1 - Make the most out of your SIMD investments
T2 - 14th International Workshop on Data Management on New Hardware, DaMoN 2018
AU - Lang, Harald
AU - Kipf, Andreas
AU - Passing, Linnea
AU - Boncz, Peter
AU - Neumann, Thomas
AU - Kemper, Alfons
N1 - Publisher Copyright:
© 2018 Copyright held by the owner/author(s). Publication rights licensed to the Association for Computing Machinery.
PY - 2018/6/11
Y1 - 2018/6/11
N2 - Increasing single instruction multiple data (SIMD) capabilities in modern hardware allows for compiling efficient data-parallel query pipelines. This means GPU-alike challenges arise: control flow divergence causes underutilization of vector-processing units. In this paper, we present efficient algorithms for the AVX-512 architecture to address this issue. These algorithms allow for fine-grained assignment of new tuples to idle SIMD lanes. Furthermore, we present strategies for their integration with compiled query pipelines without introducing inefficient memory materializations. We evaluate our approach with a high-performance geospatial join query, which shows performance improvements of up to 35%.
AB - Increasing single instruction multiple data (SIMD) capabilities in modern hardware allows for compiling efficient data-parallel query pipelines. This means GPU-alike challenges arise: control flow divergence causes underutilization of vector-processing units. In this paper, we present efficient algorithms for the AVX-512 architecture to address this issue. These algorithms allow for fine-grained assignment of new tuples to idle SIMD lanes. Furthermore, we present strategies for their integration with compiled query pipelines without introducing inefficient memory materializations. We evaluate our approach with a high-performance geospatial join query, which shows performance improvements of up to 35%.
UR - http://www.scopus.com/inward/record.url?scp=85050272441&partnerID=8YFLogxK
U2 - 10.1145/3211922.3211928
DO - 10.1145/3211922.3211928
M3 - Conference contribution
AN - SCOPUS:85050272441
T3 - 14th International Workshop on Data Management on New Hardware, DaMoN 2018
BT - 14th International Workshop on Data Management on New Hardware, DaMoN 2018
PB - Association for Computing Machinery, Inc
Y2 - 11 June 2018
ER -