D. Lea. A java fork/join framework. In ACM Java Grande 2000 Conference, San Francisco, California, June 2000. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Efficient Load Balancing for Wide-Area.. - van Nieuwpoort.. (2001) (2 citations) (Correct)
....but they are probably more complicated. Therefore, we argue that CRS is well suited for load balancing divideand conquer applications on hierarchical wide area systems. 4. PERFORMANCE EVALUATION We evaluated Satin s performance using 12 application kernels, which are taken from [7] 11] and [14], translated to Java and modified to use Satin s parallel extensions. All measurements were performed on a cluster of the Distributed ASCI Supercomputer (DAS) containing 200 MHz Pentium Pros running RedHat Linux 6.2 (kernel 2.2.16) that are connected by 1.28 Gb s Myrinet. Our results have been ....
....run only on shared memory machines and DSM systems, whereas Satin programs run on wide area systems with distributed memory. Java threads impose a large overhead, which is why Satin does not use threads at all, but uses light weight invocation records instead. The Java classes presented in [14] can also be used for divide and conquer algorithms. However, they are restricted to shared memory systems. Another Java based divide and conquer system is Atlas [3] Atlas is a set of Java classes that can be used to write divide andconquer programs. Javelin 2.0 [19] provides a set of Java ....
D. Lea. A java fork/join framework. In ACM Java Grande 2000 Conference, San Francisco, California, June 2000.
Efficient Load Balancing for Wide-Area.. - van Nieuwpoort.. (2001) (2 citations) (Correct)
....Myrinet cluster, even with wide area bandwidths of 100 KBytes s and latencies of 100 milliseconds. 5 Related Work We have discussed load balancing in Satin, a divideand conquer extension of Java. Satin has been designed for wide area systems, without shared memory. The Java classes presented in [9] can also be used for divide and conquer algorithms. However, they are restricted to shared memory systems. Most other divide and conquer systems are based on the C language. Among them, Cilk [4] only supports shared memory machines, CilkNOW [3] and DCPAR [7] run on local area, distributed memory ....
D. Lea. A java fork/join framework. In ACM Java Grande 2000 Conference, San Francisco, California, June 2000.
Satin: Efficient Parallel Divide-and-Conquer in Java - van Nieuwpoort, Kielmann, Bal (2000) (1 citation) (Correct)
....is implemented using explicit invocation records. We have implemented Satin by extending the Manta compiler. We discuss the performance of ten applications on a Myrinet based cluster. 1 Introduction There is currently much interest in divide and conquer systems for parallel programming [2, 6, 10, 11, 15]. Divide and conquer style programs start by dividing the problem into subproblems. Each subproblem is then recursively solved, again by dividing it into smaller subproblems. An example of such a system is Cilk [6] which extends C with divide and conquer primitives. Cilk runs these annotated C ....
....at eciency, Atlas was designed with heterogeneity and fault tolerance in mind, and aims only at a reasonable performance. Because Satin is compiler based, it is possible to generate code to create the invocation records, thus avoiding all runtime type inspection. The Java classes presented in [11] can also be used for divide and conquer algorithms. However, they are restricted to shared memory systems. A compiler based approach is also taken by Javar [4] In this system, the programmer uses annotations to indicate divide and conquer and other forms of parallelism. The compiler then ....
D. Lea. A java fork/join framework. In ACM Java Grande 2000 Conference, San Francisco, California, June 2000.
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