Advertisement
Published in:
2011 | OriginalPaper | Chapter
Introduction
Authors : Wolfgang Karl, Samuel Thibault, Stanimire Tomov, Taisuke Boku
Published in: Euro-Par 2011 Parallel Processing
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
The recent years have seen great research interest in exploiting GPUs and accelerators for computations, as shown by the latest TOP500 editions, whose very top entries are fully based on their use. Their potential computation power and energy consumption efficiency are appealing, but programming them however reveals to be very challenging, as not only task offloading and data transfer issues show up, but programming paradigms themselves appear to need reconsideration. Fully taping into this new kind of computation resource thus stands out as an open issue, particularly when conjointly using regular CPUs and several accelerator simultaneously. This is why we welcome this year the opening of a new “GPU and Accelerators Computing” topic along the collection of Euro-Par topics. The focus of this topic covers all areas related to accelerators: architecture, languages, compilers, libraries, runtime, debugging and profiling tools, algorithms, applications, etc.