Scaling up matrix computations on shared-memory manycore systems with 1000 CPU cores.
Fengguang SongJack J. DongarraPublished in: ICS (2014)
Keyphrases
- shared memory
- parallel architectures
- address space
- heterogeneous platforms
- compute unified device architecture
- parallel algorithm
- message passing
- multi processor
- parallel computing
- multithreading
- distributed memory
- graphics processing units
- multi core systems
- parallel architecture
- parallel programming
- graphical models
- massively parallel
- computing systems
- parallel processing
- computer systems
- data management
- distributed systems