1990
DOI: 10.1016/0743-7315(90)90031-j
|View full text |Cite
|
Sign up to set email alerts
|

The explicit token store

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
12
0
1

Year Published

1993
1993
2007
2007

Publication Types

Select...
3
2
2

Relationship

0
7

Authors

Journals

citations
Cited by 45 publications
(14 citation statements)
references
References 12 publications
0
12
0
1
Order By: Relevance
“…An array descriptor consists of four fields, viz., a data pointer to the start of S-TAM [38], and Pebbles. Unlike Pebbles, most other nonblocking models rely on variants of the Explicit Token Store architecture [6]. The *T [24] node is centered around the Motorola M88110MP with hardware support for fast context switching among threads.…”
Section: Related Workmentioning
confidence: 99%
See 2 more Smart Citations
“…An array descriptor consists of four fields, viz., a data pointer to the start of S-TAM [38], and Pebbles. Unlike Pebbles, most other nonblocking models rely on variants of the Explicit Token Store architecture [6]. The *T [24] node is centered around the Motorola M88110MP with hardware support for fast context switching among threads.…”
Section: Related Workmentioning
confidence: 99%
“…Initial values for reduction operators are generated in the appropriate threads. Threads terminate at control graph interfaces for loops and conditionals, 6 and at nodes for which the execution time is not statically determinable, such as function calls and memory (arrays and structures) accesses. These nodes are called terminal nodes.…”
Section: Top Down Code Generationmentioning
confidence: 99%
See 1 more Smart Citation
“…Several of the renements to ne-grained, static, dataow are attempts to utilize data and instruction locality [7]. In particular, hybrid dataow schemes [9] [16][17] propose a scheduling quantum larger than a single instruction in an eort to minimize the amount of synchronization while still providing an acceptable level of parallelism.…”
Section: Hybrid Dataowmentioning
confidence: 99%
“…One such application is the MIT ocean circulation model GCM [ 1 13 [2] whose data parallel implementation on the CM-5 has shown very good performance. While developing GCM for the CM-5, we also wrote a version of it in Id [ 131 for the Monsoon dataflow machine [ 141 [7]. Id is an implicitly parallel language with a multithreaded implementation model, and the GCM code allowed us to compare the In an ocean modeling code, the ocean is represented by multiple threedimensional arrays holding values for different physical characteristics of the ocean.…”
Section: Introductionmentioning
confidence: 99%