ACCESS-C Port to Cray

Notes on activities related to porting ACCESS-C1 to cray

Meeting notes from Friday 4th March 2016 =

Access-C1 build script

  • Ilia have build access-C1 executable on ngamai and passed to Wenming
  • Simple script still to be done

Testing Access-C executable

  • Wenming have also build access-C exe and qxreconf and found them identical to currently running executable
  • Identical results obtain from all 3 executable

CAWCR ACCESS-C1 test suite on ngamai

  • Xiao have passed a set of test data and results from operational run to Wenming
  • Wenming have started running access-C1 suite in CAWCR environment on ngamai
  • Results from suite is different even though main UM executable have been proven to give bit-identical result
  • Difference likely due to updates made to operational suite after it was passed to BNOC
    • VEG_FRAC from Imtiaz
    • Halo size
    • Decomposition
    • Double reconfig?

Investigating Differences

  • Xiao and Wenming to pass output results to Gary to diagnose
  • Look at difference in UM printout to identify relevant difference between BNOC and CAWCR run

ACCESS-C1 Small executables

  • Test build of APS2 small executables
  • Use small executables from APS2 in ACCESS-C1 suite where use and determine suitability


  • Lawrie and Milton looking at regridding of ACCESS-C2 output into ACCESS-C1 resolution for sending to downstream suites
    • Problem with kinks due to interpolation when using Land / Sea Mask

Last modified 6 years ago Last modified on Mar 6, 2016 9:08:35 PM