Kernel concurrency opportunities based on GPU benchmarks characterization.
Pablo CarvalhoRommel Anatoli Quintanilla CruzLúcia M. A. DrummondCristiana BentesEsteban CluaEdson CataldoLeandro A. J. MarzuloPublished in: Clust. Comput. (2020)
Keyphrases
- software systems
- real time
- kernel function
- graphics hardware
- mutual subspace method
- feature space
- kernel methods
- parallel computing
- support vector
- parallel implementation
- component analysis
- heterogeneous computing
- convolution kernel
- benchmark suite
- kernel space
- kernel regression
- kernel matrix
- machine learning
- positive definite
- kernel machines
- gpu accelerated
- reproducing kernel hilbert space
- concurrency control