Performance models and workload distribution algorithms for optimizing a hybrid CPU-GPU multifrontal solver.
Chenhan D. YuWeichung WangPublished in: Comput. Math. Appl. (2014)
Keyphrases
- learning algorithm
- machine learning algorithms
- real time
- hybrid approaches
- computational cost
- optimization problems
- graphics processors
- mathematical models
- learning models
- orders of magnitude
- theoretical analysis
- computationally efficient
- data structure
- parameter estimation
- general purpose
- benchmark datasets
- worst case
- efficient implementation
- significant improvement
- evolutionary algorithm
- graphics hardware
- feature selection