Refine
Document Type
- Article (3) (remove)
Language
- English (3)
Has Fulltext
- yes (3)
Keywords
- LOADBAL (3) (remove)
Faculty / Organisational entity
Load balancing is one of the central problems that have to be solved in parallel computation. Here, the problem of distributed, dynamic load balancing for massive parallelism is addressed. A new local method, which realizes a physical analogy to equilibrating liquids in multi-dimensional tori or hypercubes, is presented. It is especially suited for communication mechanisms with low set-up to transfer ratio occurring in tightly-coupled or SIMD systems. By successive shifting single load elements to the direct neighbors, the load is automatically transferred to lightly loaded processors. Compared to former methods, the proposed Liquid model has two main advantages. First, the task of load sharing is combined with the task of load balancing, where the former has priority. This property is valuable in many applications and important for highly dynamic load distribution. Second, the Liquid model has high efficiency. Asymptotically, it needs O(D . K . Ldiff ) load transfers to reach the balanced state in a D-dimensional torus with K processors per dimension and a maximum initial load difference of Ldiff . The Liquid model clearly outperforms an earlier load balancing approach, the nearest-neighbor-averaging. Besides a survey of related research, analytical results within a formal framework are derived. These results are validated by worst-case simulations in one-and two-dimensional tori with up to two thousand processors.
Four different initialization methods for parallel Branch-and-bound algorithms are described and compared with reference to several criteria. A formal analysis of their idle times and efficiency follows. It indicates that the efficiency of three methods depends on the branching factor of the search tree. Furthermore, the fourth method offers the best efficiency of the overall algorithm when a centralized OPEN set is used. Experimental results by a PRAM simulation support these statements.