Auto Learning and Multi-task Scheduling Framework for Heterogeneous Many-core Processors and Supercomputers
- util/ -> mpi and communcation lib
- master.c -> MPE control
- slave.c -> CPE execution
Done List:
- Parallel K-Means Clustering Algorithm with Multi-level Data-Partitioning /Liandeng Li - 2018
TODO list:
- Evulation/Metric Functions for Parameters Auto-learning
- Multi-task Resource Allocator
- Runtime Threads Scheduler
- Experimental Evaluation on Real Applications: Remote Sensing, Gene Expression.
Collaborators:
University of St Andrews:
- Teng Yu
- Dr. John Thomson
Tsinghua University/National Supercomputer Centre in Wuxi:
- Dr. Wenlai Zhao
- Prof. Haohuan Fu
- Liandeng Li
- Pan Liu
Sanger Institute:
- Dr. Shicai Wang
Publications:
- [SC18] Large-Scale Hierarchical K-means for Heterogeneous Many-core Supercomputers