Affinity-aware work-stealing for integrated CPU-GPU processors.
Naila FarooquiRajkishore BarikBrian T. LewisTatiana ShpeismanKarsten SchwanPublished in: PPOPP (2016)
Keyphrases
- graphics processing units
- multithreading
- parallel processing
- parallel computing
- gpu implementation
- heterogeneous computing
- graphics processors
- parallel computation
- general purpose
- real time
- parallel programming
- parallel algorithm
- highly parallel
- single instruction multiple data
- shared memory
- compute unified device architecture
- memory bandwidth
- distributed systems
- parallel execution
- parallel architectures
- parallel processors
- multi threaded
- highly efficient
- pc cluster
- data transfer
- massively parallel
- computational power
- efficient implementation
- graphic processing unit
- garbage collection
- parallel architecture
- high end
- graphics hardware
- processing units
- floating point
- parallel implementation