Scaling non-regular shared-memory codes by reusing custom loop schedules.
Dimitrios S. NikolopoulosErnest ArtiagaEduard AyguadéJesús LabartaPublished in: Sci. Program. (2003)
Keyphrases
- shared memory
- parallel algorithm
- message passing
- distributed memory
- parallel computing
- multi processor
- scheduling problem
- parallel programming
- learning objects
- parallel computation
- parallel machines
- parallel execution
- parallel architecture
- address space
- parallel computers
- shared memory multiprocessors
- parallel architectures
- compute unified device architecture
- parallel tree search
- multithreading
- data parallelism
- massively parallel
- belief propagation
- active contours
- search algorithm
- three dimensional