Optimal periodic remapping of dynamic bulk synchronous computations

Optimal Periodic Remapping of Bulk Synchronous Computations on Multiprogrammed Distributed Systems

For bulk synchronous computations that have nondeterministic behaviors, dynamic remapping is an effective approach to ensure parallel efficiency. There are two basic issues in remapping: when and how to remap. This paper presents a formal treatment of the first issue for dynamic computations with a priori known statistical behaviors. We have formulated the problem as two complement sequential s...

Periodic Remapping of Non-Deterministic Bulk Synchronous Computations

Dynamic remapping is critical to the performance of bulk synchronous computations that have non-deterministic behaviors because of the need of barrier synchronization between phases. This paper presents a formal analysis of the issue of when to remap for dynamic computations with a priori known statistical behaviors, with an objective of finding optimal remapping frequencies for a given toleran...

Optimal Remapping in Dynamic Bulk Synchronous Computations via a Stochastic Control Approach

A bulk synchronous computation proceeds in phases that are separated by barrier synchronization. For dynamic bulk synchronous computations that exhibit varying phase-wise computational requirements, remapping at run-time is an effective approach to ensure parallel efficiency. This paper introduces a novel remapping strategy for computations whose workload changes can be modeled as a Markov chai...

Optimal Dynamic Remapping of Parallel Computations

A large class of computations are characterized by a sequence of phases, with phase changes occurring unpredictably. We consider the decision problem regarding the remapping of workload to processors in a parallel computation when (i) Ihe uiility of remapping md the future behavior of the workload is uncertain, and (ii) phases exhibit stable execution requirements during a given phase, but requ...

Optimal Dynamic Remapping of Parallel Computations Gi Parallel C C B P G I a I X C P S F I N a L Aefort