Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ KITopen (Karlsruhe I...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
KITopen
Master thesis . 2022
Data sources: KITopen
versions View all 2 versions
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Perfect Hash Function Generation on the GPU with RecSplit

Authors: Bez, Dominik;

Perfect Hash Function Generation on the GPU with RecSplit

Abstract

Minimale perfekte Hashfunktionen (MPHFs) bilden eine statische Menge S von beliebigen Schlüsseln auf die Menge der ersten |S| natürlichen Zahlen bijektiv ab, d. h., jeder Hashwert wird exakt einmal verwendet. Sie sind in vielen Anwendungen hilfreich, zum Beispiel, um Hashtabellen mit garantiert konstanter Zugriffszeit zu implementieren. MPHFs können sehr kompakt sein — weniger als 2 Bit pro Schlüssel sind möglich. Andererseits sind MPHFs nicht in der Lage zu entscheiden, ob ein gegebener Schlüssel zu S gehört. Zurzeit ist RecSplit die speichereffizienteste MPHF. RecSplit bietet verschiedene Kompromisse zwischen Platzverbrauch, Konstruktionszeit und Anfragezeit an. RecSplit kann zum Beispiel eine MPHF mit 1.56 Bits pro Schlüssel in weniger als 2 ms pro Schlüssel konstruieren. Das ist jedoch zu langsam für große Eingaben. Diese Arbeit präsentiert neue RecSplit-Implementierungen, die Multithreading, SIMD und die Leistung von GPUs nutzen, um die Konstruktionszeit zu verbessern. Gemeinsam mit unserer neuen bijection-rotation-Methode erreichen wir Beschleunigungen um Faktoren bis zu 333 für unsere SIMD-Implementierung auf einer 8-Kern CPU und bis zu 1873 für unsere GPU-Implementierung verglichen mit der originalen, sequenziellen RecSplit-Implementierung. Dadurch können wir MPHFs mit 1.56 Bits pro Schlüssel in weniger als 1.5 μs pro Schlüssel konstruieren.

Country
Germany
Related Organizations
Keywords

ddc:004, DATA processing & computer science, info:eu-repo/classification/ddc/004, 004

  • BIP!
    Impact byBIP!
    citations
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
citations
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green