Combining CPU and GPU architectures for fast similarity search

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F12%3A10124107" target="_blank" >RIV/00216208:11320/12:10124107 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1007/s10619-012-7092-4" target="_blank" >http://dx.doi.org/10.1007/s10619-012-7092-4</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s10619-012-7092-4" target="_blank" >10.1007/s10619-012-7092-4</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Combining CPU and GPU architectures for fast similarity search

Popis výsledku v původním jazyce

The Signature Quadratic Form Distance on feature signatures represents a flexible distance-based similarity model for effective content-based multimedia retrieval. Although metric indexing approaches are able to speed up query processing by two orders ofmagnitude, their applicability to large-scale multimedia databases containing billions of images is still a challenging issue. In this paper, we propose a parallel approach that balances the utilization of CPU and many-core GPUs for efficient similaritysearch with the Signature Quadratic Form Distance. In particular, we show how to process multiple distance computations and other parts of the search procedure in parallel, achieving maximal performance of the combined CPU/GPU system. The experimental evaluation demonstrates that our approach implemented on a common workstation with 2 GPU cards outperforms traditional parallel implementation on a high-end 48-core NUMA server in terms of efficiency almost by an order of magnitude. If we

Název v anglickém jazyce

Combining CPU and GPU architectures for fast similarity search

Popis výsledku anglicky

The Signature Quadratic Form Distance on feature signatures represents a flexible distance-based similarity model for effective content-based multimedia retrieval. Although metric indexing approaches are able to speed up query processing by two orders ofmagnitude, their applicability to large-scale multimedia databases containing billions of images is still a challenging issue. In this paper, we propose a parallel approach that balances the utilization of CPU and many-core GPUs for efficient similaritysearch with the Signature Quadratic Form Distance. In particular, we show how to process multiple distance computations and other parts of the search procedure in parallel, achieving maximal performance of the combined CPU/GPU system. The experimental evaluation demonstrates that our approach implemented on a common workstation with 2 GPU cards outperforms traditional parallel implementation on a high-end 48-core NUMA server in terms of efficiency almost by an order of magnitude. If we

Druh

J_x - Nezařazeno - Článek v odborném periodiku (Jimp, Jsc a Jost)

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2012

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

DISTRIBUTED AND PARALLEL DATABASES

ISSN

0926-8782

e-ISSN

—

Svazek periodika

30

Číslo periodika v rámci svazku

3-4

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

29

Strana od-do

179-207

Kód UT WoS článku

000305520200002

EID výsledku v databázi Scopus

—