Choose timezone
Your profile timezone:
BLonD code development meeting
→
Europe/Zurich
864/1-B04 (CERN)
General Info
- New master branch done, some bug fixes and additions for speed-up
- Alex: some bug fixes do be done for the conversion of synchronous_data in preprocess
- Kostis: merge new improvements for interp_kick
- Add missing unittests!
- Failing unittests: general/cav_FB, signal processing
- Markus: push OTFB new version
- Alex: try examples with ipython notebooks, export to usual scripts
- HTML docs to be regenerated. Any docs to be added?
- Demo for writing/generating docs
- Logging: should we add log files for debugging and info
- Example for next time; demo
- Paper deadline: end of June?
- Markus: add time delay for exponential in SPS PL
BLonD++ paper
- Computer science perspective, methodological analysis
- Overall speed-up achieved from python to C++
- Extra optimisations: rf_voltage_calc(), overlapping of interp_kick() and drift() -> to come with the assembler
- Porting of phase_loop and separatrix etc. -> unittests before pushing
- SR with Intel MKL gives a huge speed-up for the RNG
- Analytical kick should be compiled with ICC if available
- pip install for /afs/, packaging
Kostis next steps
- Distributed BLonD version via MPI (uses processes) or OpenMPI (uses threads)
- Bigger test cases needed: not a problem!
- lxbatch to be tested
Windows compatibility
- Mingw not fully compatible
- TDM is old
- Update website installation guide
- Appwayer (Travis equivalent for Windows)
Markus: intermittent instabilities in simulations
- Bin size did not impact
There are minutes attached to this event.
Show them.