BLonD code development meeting

Europe/Zurich
CERN

CERN

Video conferencing only
Videoconference
BLonD code development meeting
Zoom Meeting ID
93916228605
Host
Helga Timko
Useful links
Join via phone
Zoom URL
    • 10:30 10:40
      News 10m
      • Introduction of new members
      Speakers: Alexandre Lasheen (CERN), Giulia Papotti (CERN), Heiko Damerau (CERN), Ivan Karpov (CERN), Konstantinos Iliakis (CERN), Luis Eduardo Medina Medrano (CERN), Markus Schwarz (KIT, Germany), Panagiotis Tsapatsaris (NTUA), Simon Albright (CERN), Theodoros Argyropoulos (CERN)
    • 10:40 11:30
      GPU-accelerated BLonD 50m

      Seminar and MSc Thesis summary

      Speaker: Panagiotis Tsapatsaris

      CuBLonD usage and performance

      • kick, drift, histogram, etc. comparison on single CPU/GPU with single and double precision
        • up to a factor 5 for large enough input
      • mixed CPU/GPU arrays that can be accessed and updated from either
      • multi-GPU scaling: we don't gain a factor 2 with 2 GPUs
        • some non-parallel regions and communication time
      • MPI experiments with multiple workers
        • local histogram calculation and reduce_all() to calculate global histogram
        • use of approximate methods (update only every x turns, "SRP", or assume every worker has a representative subset, "RDS")
        • 20xCPU + 1xGPU -> up to 60x speed-up with approximations
        • comparison of 2013 and 2017 GPU architecture -> good scaling, future-proof
      • Future work
        • distributed initialisation (if e.g. beam doesn't fit in a single memory)
        • OpenCL for multi-GPU
        • Experiments with equal-capability CPU & GPU -> test of load-balancing in heterogeneous HW

      Easy to merge, so it should be integrated into the basic BLonD code. Code can be used for mixed MPI multi-CPU and GPU environment. A cluster with 4 GPUs or more would be ideal. (in most heavy simulations, CPUs were used just for small calculations)