Multicore Computing

Algorithms, Architectures, and Applications
1. Auflage 2013
ISBN: 978-1-4398-5434-1
Verlag: Taylor & Francis Inc

Buch, Englisch, 452 Seiten, Format (B × H): 164 mm x 238 mm, Gewicht: 776 g

Reihe: Chapman & Hall/CRC Computer and Information Science Series

Every area of science and engineering today has to process voluminous data sets. Using exact, or even approximate, algorithms to solve intractable problems in critical areas, such as computational biology, takes time that is exponential in some of the underlying parameters. Parallel computing addresses this issue and has become affordable with the advent of multicore architectures. However, programming multicore machines is much more difficult due to oddities existing in the architectures.

Offering insights into different facets of this area, Multicore Computing: Algorithms, Architectures, and Applications focuses on the architectures, algorithms, and applications of multicore computing. It will help readers understand the intricacies of these architectures and prepare them to design efficient multicore algorithms.

Contributors at the forefront of the field cover the memory hierarchy for multicore and manycore processors, the caching strategy Flexible Set Balancing, the main features of the latest SPARC architecture specification, the Cilk and Cilk++ programming languages, the numerical software library Parallel Linear Algebra Software for Multicore Architectures (PLASMA), and the exact multipattern string matching algorithm of Aho-Corasick. They also describe the architecture and programming model of the NVIDIA Tesla GPU, discuss scheduling directed acyclic graphs onto multi/manycore processors, and evaluate design trade-offs among Intel and AMD multicore processors, IBM Cell Broadband Engine, and NVIDIA GPUs. In addition, the book explains how to design algorithms for the Cell Broadband Engine and how to use the backprojection algorithm for generating images from synthetic aperture radar data.

Zielgruppe

Researchers and graduate students in computer science and computer engineering.

Autoren/Hrsg.

Rajasekaran, Sanguthevar

Fiondella, Lance

Ahmed, Mohamed

Ammar, Reda A.

Fachgebiete

• Mathematik | Informatik EDV | Informatik Programmierung | Softwareentwicklung Programmier- und Skriptsprachen
• Mathematik | Informatik EDV | Informatik Programmierung | Softwareentwicklung Algorithmen & Datenstrukturen
• Mathematik | Informatik Mathematik Mathematik Allgemein

Inhaltsverzeichnis

Memory Hierarchy for Multicore and Manycore Processors. FSB: A Flexible Set Balancing Strategy for Last Level Caches. The SPARC Processor Architecture. The Cilk and Cilk++ Programming Languages. Multithreading in the PLASMA Library. Efficient Aho-Corasick String Matching on Emerging Multicore Architectures. Sorting on a Graphics Processing Unit (GPU). Scheduling DAG Structured Computations. Evaluating Multicore Processors and Accelerators for Dense Numerical Computations. Sorting on the Cell Broadband Engine. GPU Matrix Multiplication. Backprojection Algorithms for Multicore and GPU Architectures. Index.

Über Autor(innen)

Sanguthevar Rajasekaran is the UTC Chair Professor of Computer Science and Engineering and director of the Booth Engineering Center for Advanced Technologies at the University of Connecticut. He received a Ph.D. in computer science from Harvard University. He is a fellow of the IEEE and the AAAS and an elected member of the Connecticut Academy of Science and Engineering. His research interests include bioinformatics, parallel algorithms, data mining, randomized computing, computer simulations, and combinatorial optimization.

Lance Fiondella is an assistant professor in the Department of Electrical and Computer Engineering at the University of Massachusetts Dartmouth. He received a Ph.D. in computer science and engineering from the University of Connecticut. His research interests include algorithms, reliability engineering, and risk analysis.

Mohamed Ahmed is a program manager at Microsoft Windows Azure Mobile. He received a PhD in computer science and engineering from the University of Connecticut. His research interests include multi/many-cores technologies, high-performance computing, parallel programming, cloud computing, and GPU programming.

Reda A. Ammar is a professor and the head of the Department of Computer Science and Engineering at the University of Connecticut. He received a PhD in computer science from the University of Connecticut. He is the president of the International Society of Computers and Their Applications and editor-in-chief of the International Journal on Computers and Their Applications. His primary research interests encompass distributed and high-performance computing and real-time systems.