Year: 2011
Type: article
Abstract: Multicore computational accelerators such as Graphics Processor Units(GPUs) became common for gaining high-performance computing on a larger scale.Programming GPUs requires detailed knowledge of the underlying architecture in order to get maximum performance. In this paper we present solution of vector distance calculation on NVIDIA’s parallel computing architecture CUDA (Common Unified Device Architecture), where we optimize the performance of a parallel algorithm and get significant speedup.
Authors: Karol Grondžák, Penka Martincova, Ondrej Šuch
Language: English
Subfield: Hardware and Architecture
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2267544373