Paper Info
Title
Using stream rewriting for mapping and scheduling data flow graphs onto many-core architectures
Abstract
Dataflow graphs, consisting of concurrent actors connected by communication channels, are widely used to model multimedia applications. As dataflow graphs explicitly expose the parallelism contained in the application, they yield well to synthesis for many-core architectures. However, in case of varying and unpredictable workloads, a static mapping of actors to computing resources is often infeasible, but a dynamic approach becomes challenging due to the numerous amount of actors. Our concept of stream-rewriting represents a novel execution semantics for dataflow graphs on many-core architectures, which allows for a completely dynamic binding of actors instances to processing units. In addition, we present a distributed scheduling mechanism, global resource sharing and lightweight lock-free synchronization based on pattern matching. Also, an optimized architecture for stream-rewriting is prototyped and evaluated.
Year
DOI
Venue
2013
10.1109/ACSSC.2013.6810532
Pacific Grove, CA
Keywords
Field
DocType
multiprocessing systems,pattern matching,resource allocation,scheduling,synchronisation,communication channels,data flow graph mapping,data flow graph scheduling,distributed scheduling mechanism,execution semantics,global resource sharing,lightweight lock-free synchronization,many-core architectures,multimedia applications,pattern matching,stream rewriting
Dataflow architecture,Fair-share scheduling,Computer science,Scheduling (computing),Parallel computing,Two-level scheduling,Dataflow,Dynamic priority scheduling,Pattern matching,Data flow diagram,Distributed computing
Conference
ISSN
ISBN
Citations 
1058-6393
978-1-4799-2388-5
0
PageRank 
References 
Authors
0.34
16
4
Name
Order
Citations
PageRank
Christian Haubelt179668.77
Florian Ludwig200.34
Lars Middendorf3164.07
Christian Zebelein4385.43