Next:
Introduction
Up:
Improving the performance of
Previous:
Improving the performance of
Contents
Introduction
CP2K
HECToR
Compiling CP2K for HECToR
Benchmarks
Compiler Performance
Realspace to Planewave transfer
Halo Swapping
Non-blocking Communication
Benchmark Results
Fast Fourier Transforms
Caching to amortize costs
Transpose and MPI_Alltoallv
FFTW Planning
Load Balancing
Benchmark Results
Final Benchmarks
Further Work
Conclusion
Bibliography