Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: grouped and reordered the tasks

...

Starting list of activities/tasks:

XSEDE General:

  1. Discover available resources
  2. Access XSEDE services and resources
  3. Locate training resources to support a specific activity
  4. Get a third party application and number seeds available in any or a specific resource
  5. Get benchmark data for a given application across XSEDE resources
  6. Access Data
  7. Move Data
  8. Start a computation
  9. Terminate a computation
  10. Submit a job
  11. Parallelize existing serial code
  12. Compile a program from source on a remote machine
  13. Submit a batch job
  14. Process gigabytes of data and transfer it back to a personal machine for further analysis
  15. Extend an algorithm to handle new or larger datasets.
  16. Use profiling tools to identify bottlenecks
  17. Employ optimization techniques to improve runtime efficiency
  18. , batch or interactive
  19. Monitor job status 
  20. Get historical job information
  21. Get an estimate of start time of a submitted job

Design, Preparation:

  1. Use parallel visualization tools to analyze output
  2. Debug error messages from the compiler or linker
  3. Identify target system for a code
  4. Identify programming language for code development
  5. Identify optimized numerical libraries for use in an application code to avoid implementing a basic numerical algorithm from scratchLocate training resources to support a specific activity
  6. Use a hardware accelerator/coprocessor like: Nvidia or AMD gpu, Intel XeonPhi
  7. Find the software I need (modules?) or install the software required (system, or my $HOME) so that I can : ./configure; make

Development, measure, analysis, improvement:

  1. Parallelize existing serial code
  2. Compile a program from source on a remote machine
  3. Debug error messages from the compiler or linker
  4. Use profiling tools to identify bottlenecks
  5. Use parallel visualization tools to analyze output
  6. Process gigabytes of data and transfer it back to a personal machine for further analysis
  7. Extend an algorithm to handle new or larger datasets.
  8. Employ optimization techniques to improve runtime efficiency
  9.  Monitor job status 
  10.  Get historical job information
  11.  Get an estimate of start time of a submitted job 
  12. Get a third party application and number seeds available in any or a specific resource 
  13. Get benchmark data for a given application across XSEDE resources