Abstract
Manycore GPUs (Graphics Processing Units) are available in almost all current hardware platforms, from standard desktops to computer clusters, and thus, provide easily accessible and low-cost parallel hardware to a broad community. Originally, these processors have been designed for graphics applications; however, today there is an increasing importance for applications from scientific computing and scientific simulations. Especially for data parallel programs there can be a considerable increase of efficiency. This efficiency improvement is mainly caused by the specific hardware design of GPUs, which has been optimized for large data of graphics applications and high throughput of floating-point operations to be executed by a large number of threads. Nowadays, GPUs may comprise hundreds of cores executing these threads. A brief overview of the current architecture design of GPUs is provided in the first Sect. 7.1
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsAuthor information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Rauber, T., Rünger, G. (2013). General Purpose GPU Programming. In: Parallel Programming. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-37801-0_7
Download citation
DOI: https://doi.org/10.1007/978-3-642-37801-0_7
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-37800-3
Online ISBN: 978-3-642-37801-0
eBook Packages: Computer ScienceComputer Science (R0)