A Framework for Efficient Execution of Data Parallel Irregular Applications on Heterogeneous Systems
نویسندگان
چکیده
منابع مشابه
Efficient Parallel Execution of Irregular Recursive Programs
Programs whose parallelism stems from multiple re-cursion form an interesting subclass of parallel programs with many practical applications. The highly irregular shape of many recursion trees makes it dif-cult to obtain good load balancing with small overhead. We present a system called REAPAR that automatically parallelizes recursive C programs for SMP machines. Based on data from a single pr...
متن کاملEfficient Execution Paradigms for Parallel Heterogeneous Architectures
Koukos, K. 2016. Efficient Execution Paradigms for Parallel Heterogeneous Architectures. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1405. 54 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9654-8. This thesis proposes novel, efficient execution-paradigms for parallel heterogeneous architectures. The end of Dennard scaling is ...
متن کاملEfficient Execution of Parallel Applications in Multiprogrammed Multiprocessor Systems
Existing techniques for sharing the processing resources in multiprogrammed shared-memory multiprocessors, such as time-sharing, space-sharing, and gangscheduling, typically sacrifice the performance of individual parallel applications to improve overall system utilization. We present a new processor allocation technique that dynamically adjusts the number of processors an application is allowe...
متن کاملEfficient Execution of Parallel Java Applications
In this paper we propose mechanisms to improve the performance of parallel Java applications. The proposal is based on the establishment of a dialog between each Java application and the underlying operating system. This dialog implies modifications at the application (or compiler), the threading library and kernel levels. The paper includes some preliminary experimental results that show how a...
متن کاملFuPerMod: A Framework for Optimal Data Partitioning for Parallel Scientific Applications on Dedicated Heterogeneous HPC Platforms
Optimisation of data-parallel scientific applications for modern HPC platforms is challenging in terms of efficient use of heterogeneous hardware and software. It requires partitioning the computations in proportion to the speeds of computing devices. Implementation of data partitioning algorithms based on computation performance models is not trivial. It requires accurate and efficient benchma...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Parallel Processing Letters
سال: 2015
ISSN: 0129-6264,1793-642X
DOI: 10.1142/s0129626415500048