Skip to content

January 2014 Morning Status Meetings

sblondel edited this page Feb 6, 2018 · 1 revision

This page contains a brief summary of the status of the Xolotl development team each morning during the month of January 2014.

20140102

Crystal - Continuing to work on fitting the binding energies as functions of species size.

20140103

Crystal - Fixing a minor error in my fitting input data file and working on UQ slides.

20140106

Jay - Finishing partial derivative calculations for HeV and HeI, started re-testing tungsten_tiny.txt.
Crystal - Reviewing performance model and making remaining changes if necessary.

20140107

Jay - Trying to figure a nondimensionalization scheme for Xolotl.
Crystal - Submitting initial performance infrastructure model for review, looking at the fit some more.

20140108

Jay - Trying out a second nondimensionalization scheme to reduce the number of terms further.
Crystal - Fixing performance model problems, trying out fitting routines on simpler functions.

20140109

Jay - Figuring out what to do about characteristic concentrations in the nondimensionalized equations, looking at Crystal's performance model changes.
Crystal - Doing the fit with Python today after finding some problems yesterday in the scaling.

20140110

Jay - Looking at tungsten_tiny problems and figuring out partial derivatives for the dimensionless equations.

20140113

Jay - Emailing the PETSc about the line search bug, starting to test out tungsten_small, meetings with Brian, looking at performance model updates.
Crystal - Figuring out spline fitting coefficients, scheduling a meeting with Phil for performance infrastructure.

20140114

Jay - Looking into Mac build problems, figuring out Valgrind problem and looking at full Tungsten run.
Crystal - Sending out fit plots, looking at higher order splines in the GSL, adding tests to the model.

20140115

Jay - Running the full Tungsten problem on a cluster since the memory profiling didn't show anything bad, doing CompSci planning with Brian and David.
Crystal - Looking into higher order fits, figuring out GSL bugs.

20140116

Welcome Sophie Blondel!

Crystal - Looking into ways of determining optimal knot placement for fit, testing GPTL timer and PAPI.

20140117

Jay - Testing SuperLU_DIST in PETSc on the parallel execution bug and full tungsten problem.
Crystal - Looking at a waveform fit for the binding energies, meeting about performance infrastructure.
Sophie - Getting workstation setup, attending performance infrastructure meeting, getting a Sourceforge.net account.

20140121

Jay - Checking out possible MPICH install problems and execution bugs with full tungsten problem.
Crystal - Modeling performance infrastructure unit tests, touching up UQ slides for Sophie.
Sophie - Finishing workstation setup, getting Xolotl built.

20140122

Crystal - Writing up a description of the graphs and preparing an email for Brian, working on unit tests, touching up slides to present at lunch tomorrow for Sophie.
Sophie - Trying to build Xolotl in Eclipse, looking at the unit tests.

20140123

Jay - Trying to reproduce Barry's results with his new, super-fast and low memory version of PETSc.
Crystal - Presenting UQ slides to Sophie today, finishing testing model, looking at binding energy fit without "bad" data.
Sophie - Taking on the monumental task of trying to fix the HeInterstitial class and its tests.

20140124

Crystal - Finding good UQ references for Sophie, helping Sophie install UQTk. Using the suggestions from Tibo on the fitting problem.
Sophie - Finishing testing the flux. Starting to work with UQTk.

20140127

Jay - Meeting with Crystal and Brian about UQ stuff.
Sophie - Trying to commit the tests, looking into UQ stuff.
Crystal - Updating the performance model to use an enumeration for hardware counter types, cutting out "bad stuff" in the data and retrying the fit.

20140128

Jay - Looking into performance optimizations based on Barry's feedback. Reviewing Crystal's and Sophie's work.
Crystal - Updating the performance model to use an enumeration for hardware counter types.
Sophie - Trying to use UQTk... trying...

20140129

Crystal - Adding documentation to the UQ codes before committing it to the SVN repository.
Sophie - Working with UQ examples, cleaning Xolotl tests.

20140130

Jay - Working on iteration performance problems in getAll(). Making some slides with highlights.
Crystal - Addressing some of Jay's performance infrastructure concerns.
Sophie - Looking at Dakota examples. Learning the UML.

20140131

Jay - Working on iteration performance problems in getAll().
Sophie - Continuing to learn the UML. Starting to look at DAKOTA. Learning how to turn on PETSc monitors in Xolotl.
Crystal - Trying to get data from collaborators, working on performance infrastructure, looking at UQ fit.

Clone this wiki locally