Performance of CPU/GPU compiler directives on ISO/TTI kernels.
Sayan GhoshTerrence LiaoHenri CalandraBarbara M. ChapmanPublished in: Computing (2014)
Keyphrases
- graphics processing units
- general purpose
- graphics processors
- gpu implementation
- level parallelism
- memory bandwidth
- parallel programming
- heterogeneous computing
- graphics hardware
- real time
- feature space
- programming language
- kernel function
- parallel implementation
- parallel computing
- cpu implementation
- parallel processing
- graphic processing unit
- support vector
- power saving
- parallel computation
- multiple kernel learning
- kernel methods
- gpu accelerated
- data transfer
- compute unified device architecture
- massively parallel
- computing systems
- distributed memory machines
- multi threaded
- multithreading
- kernel learning
- java virtual machine
- highly optimized
- instruction scheduling
- intel xeon