Benchmarking

Our benchmarking

1. OpenMP vs MPI on Beowolf Cluster OpenMP vs MPI on Beowolf Cluster
2. VASP performance data
3. Bperformance of F90 TBMD code for netals
4. Performance data using a fortran code
   • 21 June 1999, Parallel scaling on Origin2000
   • 6 Feb 1998 data, Origin, T3E
   • 16 Dec 1997 data, Origin, T3E, Exemplar
   • Larry Smarr's Alliance'98 Performance on Origin2000 data.
   • NCSA performance highlights; read at your own risk.
5. Performance data using c++ code
   • single-processors
6. Performance data for Standard Template Classes (C++)
   • data
   • benchmark code

Other Benchmarking

• IDC November 01, new HPC ranking method
• ORNL Benchmarking
  • ORNL Netlib
  • Many benchmarks
  • Linpack performance on nearly everything (postscript file)
• Review of Analysis Performance Tools for MPI code, Brown, Dongarra, London