Alice Weekly Meeting: Software for Hardware Accelerators

Wednesday 14 Oct 2026, 10:00 → 11:30 Europe/Zurich

10:00 → 10:20 Discussion

Speaker: David Rohr (CERN)

10:20 → 10:25 TPC ML Clustering

Speaker: Christian Sonnabend (CERN)

BrevitasWorkshop_19102026.pdf

Neural network seeding vector (https://github.com/AliceO2Group/AliceO2/pull/15756)

• Found a bug in the implementation (missing setting during application, seeding vectors were never used) -> Fixed
• Results don't change much. 0-100% centrality is mostly unchanged, but one improvement can be seen in 0-5% centrality enforced simulation

 

Is it worth the complications? 🤷‍♂️

 

ROOT: TMVA SOFIE inference of .onnx models on the EPNs

• Discussion during ROOT Q3 meeting last week: ALICE has great interest to give SOFIE a try
• One night of synergetic work between me and ChatGPT Codex: Huston, we have liftoff!
  • ROOT PR: https://github.com/root-project/root/pull/23657 (required as dev package for epn build)
  • O2 PR: https://github.com/AliceO2Group/AliceO2/pull/15913
  • alienist PR: https://github.com/alisw/alidist/pull/6369

 

This brings:

• Support for cuda_ / hip_stream
• GPU code generation and on-the-fly compilation of .onnx models. Inflight-CCDB-loaded models are supported
• Support for FP16 and FP32
• Support for dense layers and ReLU operations
• Support for memory arena pointer (provided by reco framework as scratch allocation)
• Direct interface with SOFIE's generalised kernel generation code

 

Brevitas presentation

• Slides attached, please have a look
• Presentation is next Monday at 14:30

 

10:25 → 10:30 GPU Parameter Optimizations

Speaker: Gabriele Cimador (CERN, Università and INFN Torino)

10:30 → 10:35 Efficient Data Structures

Speaker: Dr Oliver Gregor Rietmann (CERN)

 

Implement NGT SoA Code in O2 standalone benchmark

• Working on this fork of the AliceO2 repo, with a CI pipeline:
  • Running on NGT hardware with 4 different GPUs (Nvidia and AMD)
  • Extended CI-pipline to fail if GPU.out changes
• Implemented SoA in SectorTracker: GPUTPCBaseTrackParam, GPUTPCTrackParam, GPUTPCTracklet, GPUTPCTrack

 

• Make better use of SoA to improve performance
  • ExtrapolationTracking
    • Register spill from rowHits array, but the accesses are not coalesced.
    • global memory loads on average use only 4.4 out of 32 bytes sector.
    • active warps per thread changed from 2.65 out of 32 (8%) to 
    • I made TrackletSelector filter the Tracklets that cross sector boundaries, so that we extrapolate only those.
    • Sorting extrapolation candidates by estimated length (row where it leaves the sector): Made kernel duration 18% faster on average.

 

metric	old	new	newer
active warps per thread	3 / 32	15 / 32	18 / 32
average SM occupancy	26%	10%	9.38%
bytes utilized per loaded sector	4.4 / 32	12 / 32	11.6 / 32
register spill to lcoal memory (bytes)	1,402,723	19,308	18,003

 

• Next Steps
  • Put sorting for TrackletSelector and ExtrapolationTracking together and see if we can avoid code dublication.
  • Try AoS vs SoA again.
  • Generate GPU parameters .csv file for the newest changes.

10:40 → 10:45 TPC Clusterization / OpenCL / Highly Ionizing Particles

Speaker: Felix Weiglhofer (CERN)

10:45 → 10:50 ITS Tracking

Speakers: Felix Schlepper (CERN, Heidelberg University (DE)), Gabriele Cimador (CERN, Università and INFN Torino), Matteo Concas (CERN)

10:50 → 10:55 System Run Coordination Topics

Speaker: Ernst Hellbar (CERN)