ALICE O2 CI-Pipelines on NGT Cluster

• A fork of the AliceO2 repo is not in the  NextGenTrigggers (NGT) GitHub organization
• It has a GitHub action running the standalone benchmark on NGT GPUs
• Uses the builds in /cvmfs/alice.cern.ch/ of O2 and dependencies (pipeline takes about 7 minutes)
• Different GPUs are tested in parallel on different VMs
• O2 standalone benchmark works on the all ngt-resources:
  
  • Nvidia H100 188GB NVL
  • AMD Instinct MI300X
  • AMD Radeon Pro W7900
  • Nvidia L40S
    
• We are now using custom .par files
• Next steps
  • Generate optimized .par files with Gabriele
• Possible next steps
  • Add new architectures to O2 hardcoded ones?
  • Store results in csv format and plot them?
  • Display the plot in the github web gui?

Implement NGT SoA Code in O2 standalone benchmark

• Working on this fork of the AliceO2 repo
• Simplified the AoS-SoA wrapper code
• Started to apply our AoS-SoA code to: 
  • Merger
    • GPUTPCGMSectorTrack
    • GPUTPCGMTrackParam 
  • SectorTracker
    • GPUTPCBaseTrackParam
    • GPUTPCTrackParam
    • GPUTPCTrack
• Started with SectorTracker, no problems until now

Preparation for ACAT Conference (8. - 12. September)

• Presenting with Jolly our joint work on AoS/SoA
  • Jolly: SoA using reflection (C+++26)
  • Oliver: SoA using templates (C++17)
• We will present the benchmark results comparing different AoS/SoA implementations
• Fixed some issues with a experimental compiler that supports reflections
• Problems:
  • Example where AoS is faster than SoA on CPU?
  • Sometimes loops over SoA are not vectorized with the new experimental compiler for reflections