- Titel
- Conflict Detection-Based Run-Length Encoding
- Untertitel
- AVX-512 CD Instruction Set in Action
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-821799
- Konferenz
- International Conference on Data Engineering Workshops. Paris, 16.-20.04.2018
- Quellenangabe
- 2018 IEEE 34th International Conference on Data Engineering Workshops (ICDEW)
Erscheinungsort: New York, NY
Verlag: IEEE
Erscheinungsjahr: 2018
Seiten: 96-101 - Erstveröffentlichung
- 2018
- Abstract (EN)
- Data as well as hardware characteristics are two key aspects for efficient data management. This holds in particular for the field of in-memory data processing. Aside from increasing main memory capacities, efficient in-memory processing benefits from novel processing concepts based on lightweight compressed data. Thus, an active research field deals with the adaptation of new hardware features such as vectorization using SIMD instructions to speedup lightweight data compression algorithms. Following this trend, we propose a novel approach for run-length encoding, a well-known and often applied lightweight compression technique. Our novel approach is based on newly introduced conflict detection (CD) instructions in Intel's AVX-512 instruction set extension. As we are going to show, our CD-based approach has unique properties and outperforms the state-of-the-art RLE approach for data sets with small run lengths.
- Andere Ausgabe
- Link zum Artikel, der zuerst in der IEEE Xplore Digital Library erschienen ist
DOI: 10.1109/ICDEW.2018.00023 - Freie Schlagwörter (DE)
- Leichte Datenkompression, Vektorisierung, AVX 512, Lauflängenkodierung
- Freie Schlagwörter (EN)
- Lightweight Data Compression, Vectorization, AVX 512, Run Length Encoding
- Klassifikation (DDC)
- 005
- Verlag
- IEEE, New York, NY
- Förder- / Projektangaben
- Deutsche Forschungsgemeinschaft (DFG)
Collaborative Research Center 912
HAEC - Highly Adaptive Energy-Efficient Computing
(SFB 912)
ID: 164481002 - Version / Begutachtungsstatus
- angenommene Version / Postprint / Autorenversion
- URN Qucosa
- urn:nbn:de:bsz:14-qucosa2-821799
- Veröffentlichungsdatum Qucosa
- 18.01.2023
- Dokumenttyp
- Konferenzbeitrag
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
