Elimination Techniques of Redundant Data Transfers Among GPUs and CPU on Recursive Stream-Based Applications.
Pablo Lamilla ÁlvarezShinichi YamagiwaMasahiro AraiKoichi WadaPublished in: IPDPS Workshops (2011)
Keyphrases
- redundant data
- graphics processing units
- gpu implementation
- general purpose
- graphics processors
- data streams
- real time
- sliding window
- parallel processing
- heterogeneous computing
- graphics hardware
- parallel programming
- neural network
- massively parallel
- computing systems
- processing units
- stream data
- memory bandwidth
- learning algorithm
- efficient implementation
- recursive queries
- dead end
- compute intensive
- database systems
- multi stream
- computational power
- parallel architectures
- personal computer
- streaming data
- parallel computing