Design Flow for Reconfiguration Based on the Overlaying Concept

Meisel, A.; Draeger, A.; Schneider, Sven; Hardt, Wolf‐Dietrich

doi:10.1109/rsp.2008.20

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Managed resource types might be processor type, speed, or version for atomic processing units, or latency and bandwidth for communication links. But resource models can be much more complex in order to describe reconfigurable systems that are prepared to perform a number of specific tasks as described in [6]. Even communication links could dynamically adapt to different combinations of bandwidths and latencies, using techniques for interface reconfiguration [7].…”

Section: Hierarchical Data Structuresmentioning

confidence: 99%

Communication-Aware Hierarchical Online-Placement in Heterogeneous Reconfigurable Systems

Schneider

Meisel

Hardt

2009

2009 IEEE/IFIP International Symposium on Rapid System Prototyping

Self Cite

View full text Add to dashboard Cite

In recent years, there was a tendency towards heterogeneous systems. The main property of these systems is that communication resources and computation resources are non-uniform in architecture and performance. Often, communication links between physically close infrastructure nodes provide high performance, e.g. in the sense of low latency or high bandwidth. A placement methodology for such systems should minimise the overall communication costs and prevent blocking of long, low-performanced communication links. Therefore application components with higher communication demands should be located closer to each other. For this, we propose a placement methodology that is based on a hierarchical application and infrastructure description. Due to our hierarchical approach, no knowledge about the potentially complex semantics of multi-dimensional communication demands is needed for the optimisation during the run-time of the placement. Our implementation shows a simple but efficient onlineplacement which also leads to a higher robustness in distributed embedded systems.

show abstract

Section: Hierarchical Data Structuresmentioning

confidence: 99%