News
- Migrated website: videos missing
- Issue on BLonD forum
- Answered by Simon and Alex
Update Kostis
- With Theo, simulations on MPI cluster
- Scripts to generate input file in SLURM cluster (based on scripts for HTcondor)
- Smooth histogram (test cases of Ivan)
- Theo's suggestions: split beam statistics to dE, dt statistics OK
- H-BLonD paper in Computing Frontiers conference got best paper award; Congratulations!!
GPU-BLonD (Panagiotis)
- See README for instructions
- Installation: CUDA, PyCUDA, Scikit-CUDA
- Usage instructions bm_use_gpu(): initialise and update functions
- Instructions for lxplus
- Example case (Simon) minimalWorking.py in repo
- Use cuprof for profiling
- Faster and good accuracy
- Supported: tracker, profile, ... not all functions
- Access GPU on CERN: quite some waiting time, but only on cluster
- Submit file to be uploaded
Present status of Ring and RingSection (Alex)
- Ring works with list of Section in Alex's fork
- Some examples and unittests available
- Docs upload automatically upon merge of pull request (to be propagated to BLonD)
Orbit in EOM (Helga)
- Defines reference clock
- Defines referene energy from B-field
- Drift equation contains individual orbit/radius
Renaming (all)
- input_parameters -> machine_parameters
- Ring, RingSection
- keep Ring.circumference
- Ring.length, RingSection.length
- length and length_design?
- length_design should be defined from energy reference, and defines also RF design frequency and phase
- length should introduce a phase shift in actual RF phase
- orbit_bump for offset from length_design
- Document well *_d definition!
- Alpha_orders: list of orders used
- etas not used anymore, but keep option to derive
- alpha_orders has a finite length, e.g. 10
Next time: discuss pull request with Ring and RingSection
There are minutes attached to this event.
Show them.