• --: check the pAMG wiki: https://dis.cs.uiuc.edu/trac/pamg/timeline
• --: Lookup the documentation on what's needed to make a krylov solver
• --: Implement a dummy krylov solver
• --: modify an example to implement this krylov solver
• --: Write something to input matrix market, fdiff, and FEM matrix routines into Petsc.
• --: email the PETSc developers and tell them that BLACS wont compile on my home machine.
• email Luke, ask him for smooth aggregation papers.
• schedule 1 hour of pAMG time.
• --: Email back David with a better explanation of the errors I've found in CJLP
• --: make some experiments showing poor parallelization.
• 2008-04-12: Test the code on 3 processors using my machine.

• --: Finish solver:
  • change build strength matrix so it returns the above and below matrices
  • change the interpolation routines so that they use these new matrices
  • move the creation of the sets inside the interpolation routines.
• --: Write a report.

--: +run a parallel timing test on turing (if possible)

Reference

https://dis.cs.uiuc.edu/trac/pamg/timeline

http://www-unix.mcs.anl.gov/petsc/petsc-as/documentation/index.html