Advancing the distributed Multi-GPU ChASE library through algorithm optimization and NCCL library.
Xinzhe WuEdoardo Di NapoliPublished in: SC Workshops (2023)
Keyphrases
- optimization algorithm
- experimental evaluation
- dynamic programming
- matching algorithm
- preprocessing
- learning algorithm
- detection algorithm
- computational cost
- np hard
- k means
- computational complexity
- objective function
- bi level
- real time
- expectation maximization
- segmentation algorithm
- times faster
- combinatorial optimization
- optimization model
- constrained optimization
- multi robot
- global optimization
- gpu implementation
- linear programming
- simulated annealing
- high accuracy
- significant improvement
- search space
- relational databases