Improving CPU Performance Through Dynamic GPU Access Throttling in CPU-GPU Heterogeneous Processors.
Siddharth RaiMainak ChaudhuriPublished in: IPDPS Workshops (2017)
Keyphrases
- graphics processing units
- heterogeneous computing
- gpu implementation
- graphics processors
- general purpose
- parallel computing
- parallel processing
- multithreading
- parallel computation
- memory bandwidth
- computing systems
- real time
- graphics hardware
- parallel algorithm
- parallel implementation
- compute unified device architecture
- processing units
- parallel programming
- high end
- high performance computing
- massively parallel
- cpu implementation
- parallel architectures
- highly parallel
- floating point
- intel xeon
- efficient implementation
- pc cluster
- commodity hardware
- compute intensive
- data transfer
- graphic processing unit
- access control
- single instruction multiple data
- dynamic environments
- memory hierarchy
- personal computer