Improved MPI collectives for MPI processes in shared address spaces.
Shigang LiTorsten HoeflerChungjin HuMarc SnirPublished in: Clust. Comput. (2014)
Keyphrases
- parallel implementation
- parallel algorithm
- message passing
- message passing interface
- general purpose
- high performance computing
- shared memory
- parallelization strategy
- parallel computing
- bayesian networks
- massively parallel
- distributed memory
- improved algorithm
- parallel computers
- collective learning
- collective intelligence
- process model
- database systems