Fast and scalable parallel matrix computations with reconfigurable pipelined optical buses.
Keqin LiPublished in: Parallel Algorithms Appl. (2004)
Keyphrases
- parallel computation
- parallel architecture
- parallel implementation
- fine grain
- low cost
- distributed memory
- parallel processing
- systolic array
- coarse grain
- linear algebra
- matrix multiplication
- general purpose
- hardware implementation
- map reduce
- highly scalable
- efficient implementation
- parallel programming
- singular value decomposition
- reconfigurable architecture
- massively parallel
- computer architecture
- functional units