May 27 2019 » History » Version 3
May 27 2019¶
Mark (FE workshop Göttigen), Berk (parental leave)
Completed since last meeting (May 20 2019)¶
- A bit more code integrated.
- Broken virial in SHAKE #2879 - fixed with follow-up from Berk.
- Mysterious glibc/gcc/gomp/avx512 failures #2762 turned out to be a compiler bug. Fixed.
- BioExcel All Hands Munich (20-21 May).
- Rewrite of checkpoint and restart logic.
- EnergyOutput refactoring.
- 2018.7 this week - see https://redmine.gromacs.org/versions/90
- Post-submit failing in complex.nbnxn-ljpme-LB-geometric (issue #2939) - Issue identified.
- GPU bonded optimization (issue #2818, 'AMBER trick')
- Meeting with Intel
- Half year plan meeting to discuss what should be attempted to merge into 2020 in June.
Progress since last meeting¶
- Cryo EM finished by end of the week - Christian
- BioExcel-2 MS6.2 user survey planning on track - Christian
- BioExcel-2 CSC workshops, all set up, applied for PRACE funding - Christian
- Extracting MPI setup from commrec setup, patches - needs review
- WIP of Docker+k8s documentation build, going forward but killed Marks PC - Mark
- PRACE Seasonal School June 10-13 - Erik, Rossen
- BioExcel-2 Sardinia summer school July - Christian & Mark
- CUDA update/constrain patches, module needs unification - Artem
- CUDA buffer ops patches - Szilard, Artem
- Constant pH code - working branch, code not very clean, nothing new for now - Berk, Paul
- Topology data structure cleanup - Paul, Mark
- Trajectory analysis output filters - Paul, Mark