Efficient Distributed-Memory Parallel Matrix-Vector Multiplication with Wide or Tall Unstructured Sparse Matrices.
Jonathan EcksteinGyorgy MatyasfalviPublished in: CoRR (2018)
Keyphrases
- distributed memory
- sparse matrices
- matrix multiplication
- floating point
- shared memory
- linear algebra
- rows and columns
- ibm sp
- parallel computers
- parallel implementation
- data parallelism
- condition number
- multiprocessor systems
- computer architecture
- message passing
- parallel processing
- parallel machines
- sparse matrix
- image processing
- multistage
- data processing