Accelerating MPI Collective Communications through Hierarchical Algorithms Without Sacrificing Inter-Node Communication Flexibility.
Benjamin S. ParsonsVijay S. PaiPublished in: IPDPS (2014)
Keyphrases
- neural network
- communication systems
- search algorithm
- computational complexity
- orders of magnitude
- computational cost
- general purpose
- optimization problems
- communication patterns
- data sets
- parallel architectures
- benchmark datasets
- computationally efficient
- worst case
- significant improvement
- learning algorithm
- genetic algorithm
- data mining