September 2020 Morning Status Meetings - ORNL-Fusion/xolotl GitHub Wiki

This page contains a brief summary of the status of the Xolotl development team each morning during the month of September 2020.

20200901

Sophie - Trying PETSc reduced matrix method.

Sophie - Continuing on the reduced matrix method and checking analysis scripts in the plsm branch.

Sophie - Updating Cuda and checking Xolotl with C++17.

Sophie - Looking at what is wrong with the reduced matrix method. Continuing to update Cuda.

Sophie - Continuing to look at what is wrong with the reduced matrix method. Comparing Cori and Summit performances.

Sophie - Starting to profile 3D communication for NE coupling code. Gathering dynamic grid data.

Sophie - Continuing on the reduced matrix method.

Sophie - Testing Cori affinity and fixing the Fe network in plsm branch.

Sophie - Looking at the number of PETSc vectors in Xolotl and timing the NE case.

Sophie - Testing CUDA 11 on Iguazu and fixing the use of initGBLocation() in plsm branch.

Sophie - Profiling the reduce matrix cases and looking at where all the Vec are used.

Sophie - Continuing on the reduced matrix method and testing different grids for PSI.

Sophie - Verifying all PSI network build correctly. Changing some of the option names.

Sophie - Implementing the reduced matrix method permanently.

Sophie - Profiling the plsm-network branch with the geometric gradation grid.

Sophie - Continuing to test the geometric gradation grid. Working on grouping for the Fe case.