Suggestion for Evolutionary Strategy Implementation in CUDA

Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21260%2F10%3A00171990" target="_blank" >RIV/68407700:21260/10:00171990 - isvavai.cz</a>
Výsledek na webu
—
DOI - Digital Object Identifier
—

Jazyk výsledku
angličtina
Název v původním jazyce
Suggestion for Evolutionary Strategy Implementation in CUDA
Popis výsledku v původním jazyce
Nowadays, there exist cheap source of significant computation power - GPGPU (General Purpose computation on Graphics Processing Unit). Unfortunatelly, many implementations of GAs (Genetic Algorithms) and Evolutionary Strategies on GPGPUs follow implementations on general purpose CPUs and they are not so efficient as they might be. The paper discusses differences between standard CPUs, parallel systems based on CPUs (e.g. multi-kernel CPUs) and GPGPUs. Then, the main approaches to GPGPU programming are studied and novel implementation of Evolutionary Strategy is presented.
Název v anglickém jazyce
Suggestion for Evolutionary Strategy Implementation in CUDA
Popis výsledku anglicky
Nowadays, there exist cheap source of significant computation power - GPGPU (General Purpose computation on Graphics Processing Unit). Unfortunatelly, many implementations of GAs (Genetic Algorithms) and Evolutionary Strategies on GPGPUs follow implementations on general purpose CPUs and they are not so efficient as they might be. The paper discusses differences between standard CPUs, parallel systems based on CPUs (e.g. multi-kernel CPUs) and GPGPUs. Then, the main approaches to GPGPU programming are studied and novel implementation of Evolutionary Strategy is presented.

Rok uplatnění
2010
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Podobné výsledky(10)