A common machine language for grid-based architectures

Thies, William; Karczmarek, Michal; Gordon, Michael I.; Maze, David; Wong, Jeremy; Hoffmann, Henry; Brown, Matthew G.; Amarasinghe, Saman

doi:10.1145/571666.571673

Cited by 3 publications

(2 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use the StreamIT compiler [16], [17] and the set of stream applications distributed by the StreamIT project (listed in the first column of Table 2). We use the StreamIT compiler's built-in functions to create loadbalanced sets of actors that are then laid-out on our networks (4x4 and 6x6 mesh networks).…”

Section: Methodsmentioning

confidence: 99%

Disjoint-path routing: Efficient communication for streaming applications

Seo

Thottethodi

2009

2009 IEEE International Symposium on Parallel &Amp; Distributed Processing

View full text Add to dashboard Cite

Streaming is emerging as an important programming model for multicores. Streaming provides an elegant way to express task decomposition and inter-task communication, while hiding laborious orchestration details such as load balancing, assignment (of stream computation to nodes) and computation/communication scheduling from the programmer. This paper develops a novel communication optimization for streaming applications based on the observation that streaming computations typically involve large, systematic data transfers between known communicating pairs of nodes over extended periods of time. From the above observation, we advocate a family of routing algorithms that expend some over overheads to compute disjoint paths for stream communication. Disjoint-path routing is an attractive design point because (a) the overheads of discovering disjoint paths are amortized over large periods of time and (b) the benefits of disjoint path routing are significant for bandwidth-sensitive streaming applications. We develop one instance of disjoint-path routing called tentacle routing -a backtracking, besteffort technique. On a 4x4 (6x6) system, tentacle routing results in 55% (84%) and 28% (41%) mean throughput improvement for high-network-contention streaming applications, and for all streaming applications, respectively.To motivate our design, we begin by describing the two key inefficiencies of coherence networks in supporting streaming communication. The first inefficiency is the mismatch between "pull-based" communication supported by coherence networks and the "push-based" communication used by streaming applications. To understand this mismatch, consider the least disruptive way to support streaming on coherence networks of shared memory multicores. In such a design, streaming would be implemented using the underlying shared memory (i.e., by implementing streams as softwarebased FIFOs) and no hardware changes would be needed. Coherent shared memory supports pull based communication where the producing actor deposits new values in its local caches, which are then "pulled" by coherence mechanisms by consuming actors. In contrast, stream communication is inherently push-based wherein producing actors actively push data to consuming actors. Supporting push-based communication using pull-based mechanisms wastes bandwidth because pull-based communication requires a request and a response, rather than a simple data packet. In addition, because pull-based communication requires producerconsumer synchronization, there may be other packets to be exchanged as well. Traffic : S {C1, C2, C3, C4} {C1, C2, C3, C4} J Traffic : S {C1, C2, C3, C4} {C1, C2, C3, C4} J S J C1 C2 C3 C4 S J C1 C2 C3 C4 A) Original Graph & 4x4 MESH Mapping C1 C2 C3 C4 J S Mapping Contention C1 C4 C2 C3 S J C) Disjoint Path Routing B) DOR

show abstract

Section: Methodsmentioning

confidence: 99%

Disjoint-path routing: Efficient communication for streaming applications

Seo

Thottethodi

2009

2009 IEEE International Symposium on Parallel &Amp; Distributed Processing

View full text Add to dashboard Cite

show abstract

“…However, Spidle is oriented towards synchronous, single-process applications [9], and StreamIt aims for streaming DSP applications and processors with grid based architectures [10], [11].…”

Section: Related Workmentioning

confidence: 99%

Code generation for WSLAs using AXpect

Swint

2004

Proceedings. IEEE International Conference on Web Services, 2004.

View full text Add to dashboard Cite

Abstract-WSLAs can be viewed as describing the service aspect of web services. By their nature, web services are distributed. Therefore, integrating support code into a web service application is potentially costly and error prone. Viewed from this AOP perspective, then, we present a method for integrating WSLAs into code generation using the AXpect weaver, the AOP technology for Infopipes. This helps to localize the code physically and therefore increase the eventual maintainability and enhance the reuse of the WSLA code. We then illustrate the weavers capability by using a WSLA document to codify constraints and metrics for a streaming image application that requires CPU resource monitoring.

show abstract