Probabilistic analysis of the optimal efficiency of the multi-level dynamic load balancing scheme: (extended abstract)

Abstract: "This paper investigates the optimal efficiency of the multi-level dynamic load balancing scheme for OR-parallel programs, using probability theory. In the single-level dynamic load balancing scheme, one processor divides the given task into several subtasks, which are distributed to...

Full description

Saved in:
Bibliographic Details
Main Authors: Kimura, Kouichi (Author), Ichiyoshi, Nobuyuki (Author)
Format: Book
Language:English
Published: Tokyo, Japan 1990
Series:Shin-Sedai-Konpyūta-Gijutsu-Kaihatsu-Kikō <Tōkyō>: ICOT technical memorandum 974
Subjects:
Summary:Abstract: "This paper investigates the optimal efficiency of the multi-level dynamic load balancing scheme for OR-parallel programs, using probability theory. In the single-level dynamic load balancing scheme, one processor divides the given task into several subtasks, which are distributed to the other processors on demand and then executed independently. We introduce a formal model of the execution as a queuing system with several servers. And we investigate the optimal granularity of the subtasks to attain the maximal efficiency, taking account of the dividing costs and the load imbalance among the processors. Thus we obtain the estimates of the maximal efficiency
We then apply these results to the analysis of the efficiency of the multi-level dynamic load balancing scheme, which is the iterated application of the single-level dynamic load balancing scheme in a hierarchical manner. And we show how the scalability of the load balancing scheme is thereby improved over the single-level one.
Physical Description:5 S.

There is no print copy available.

Interlibrary loan Place Request Caution: Not in THWS collection!