R. Jagannathan, C. Dodd, and I. Agi. GLU: A high-level system for granular data-parallel programming. Concurrency: Practice and Experience, (1):63-83, 1997. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Models of Computation for Specification and Design Methodology.. - Janka (1999) (Correct)
....have been defined to handle different types of models and mappings, typically trading expressiveness for formal properties. They include Karp and Miller s computation graphs [12] Lee and Messerschmitt s synchronous DFG s [13] Kaplan et al. s processing graph method (PGM) 14] and Granular Lucid [15]. Of particular interest is synchronous dataflow (SDF) 13] which requires processes to consume and produce a fixed number of tokens for each firing. The SDF MoC has the useful property that a finite static schedule can always be found that will return the graph to its original state, where state ....
R. Jagannathan, C. Dodd, and I. Agi, "GLU: A High-Level System for Granular Data-Parallel Programming, " Concurrency: Practice and Experience, vol. 9, pp. 63-83, 1997.
Task System and Item Architecture (TSIA) - Steinmacher-Burow (1999) (Correct)
....help relate this presentation to the presentations of precursors. 2. 2 A Task is Not a Process For an application which executes in terms of tasks, many projects have demonstrated that a task system (TS) can provide such an application with execution elements [168 E] CilkNOW ] DBC] DNA] Funnel] GLU] Internet Computing] Jade] Linda Piranha] Mentat] Nimrod] Packet Filter] RTU] SCHEDULE] SETI home] Spawn] SUMO] Supervisor Worker] From these previous TS, TSIA adopts and unifies the notion of tasks and the role of the TS. As described in section 2.11, the combined experiences of ....
....between tasks must be known to the TS, since the tasks must execute in a sequence which satisfies the dependencies. For example, the dependencies arising from the application definition may be specified explicitly to the TS [Cilk NOW] Jade] SCHEDULE] or be automatically recognized by the TS [GLU] Mentat] Of course, independent tasks may execute in parallel. For example, two tasks are independent if the items of one task are independent of the items of the other task. Of course, identical in effectively may be independent if the in may be copied or otherwise simultaneously accessed. ....
[Article contains additional citation context not shown here]
R. Jagannathan, C. Dodd and I. Agi. "GLU: A High-Level System for Granular DataParallel Programming", Concurrency: Practice and Experience, 1996 (to appear). (This and other GLU information is available at http://www.csl.sri.com/GLU.html). 238
WebComs: Transactions as Object-Flow Networks for the WOS - Schubiger, Krone.. (1999) (1 citation) (Correct)
....System, Flow Based Programming, Coordination, Distributed Systems, Concurrency, Communication Models. 1 Introduction Data flow networks are widely recognized as a suitable coordination language for distributed applications. Their advantages that is easy to understand diagrams, natural parallelism [5] and fault tolerance [3] are the result of research being conducted in different areas such as programming languages [16] 2] coordination languages [4] and programming paradigms [12] Data flow networks have several advantages in a distributed context: ffl Data flow networks exhibit a natural ....
....Upon failure of either a channel or a component, the network can be automatically reconfigured to use alternative components and channels. ffl Data flow networks have a wide coverage in the literature. Entire languages and systems are build on the data flow paradigm [12] such as Lucid [16] GLU, [5], Strand [2] just to mention a few. 1 ffl Data flow networks are used successfully in many real world settings. Applications range from hardware (super scalar pipelined CPU s) to tools (development environments) to business applications (online transaction systems) Using a data flow network in ....
R. Jagannathan, C. Dodd, and I. Agi. GLU: A High-Level System for Granular DataParallel Programming. October 1995.
GLU Programmer's Guide - Version 1.0 - Jagannathan, Dodd (1996) Self-citation (Jagannathan Dodd) (Correct)
....and heterogeneous networks as ubiquitous parallel platforms. The programmer must address how each program that is written will adapt to various dynamics, including changing load and availability of computing resources as well as network resources. 1. 2 GLU Approach to Parallel Programming GLU [3] is a software system for rapidly constructing portable, efficient, and adaptable parallel programs, either from scratch or by reusing existing sequential code. The essential idea behind GLU is that a parallel application can be described as an implicitly parallel graph composition of external, ....
R. Jagannathan, C. Dodd, and I. Agi. GLU: A high-level system for granular dataparallel programming, Concurrency: Practice and Experience, John Wiley and Sons, 1997. (Preprint in http://www.csl.sri.com/glu/html/papers.)
GLU Implementation Architectures for Heterogeneous Systems - Jagannathan, Dodd Self-citation (Jagannathan Dodd) (Correct)
No context found.
R. Jagannathan, C. Dodd, and I. Agi. GLU: A high-level system for granular data-parallel programming. Concurrency: Practice and Experience, 1996.
The Semantics Of Dimensions As Values Joey Paquet - Paquet, Plaice (1999) (Correct)
No context found.
R. Jagannathan, C. Dodd, and I. Agi. GLU: A high-level system for granular data-parallel programming. Concurrency: Practice and Experience, (1):63-83, 1997.
The GIPSY Architecture - Paquet, Kropf (Correct)
No context found.
R. Jagannathan, C. Dodd, and I. Agi. GLU: A high-level system for granular data-parallel programming. Concurrency: Practice and Experience, (1):63--83, 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC