Punit Sharma (BNL): ATLAS EF Tracking for HL-LHC

Monday 4 May 2026, 12:45 → 14:00 US/Eastern

D-122 (SBU Physics building)

Elizabeth Worcester, Valerio Dao (Stony Brook University)

12:45 → 13:30 ATLAS EF Tracking for HL-LHC

The High-Luminosity LHC presents an unprecedented challenge for the ATLAS experiment, with instantaneous luminosities generating up to 200 simultaneous collisions (μ = 200). To maintain high trigger efficiency for crucial physics signatures like b-hadrons and tau-leptons, rapid track reconstruction within the Event Filter is strictly required. Initially reliant on custom built silicon, ATLAS made a strategic pivot to commercial heterogeneous computing with EFTracking. This seminar explores the multi-year R&D effort evaluating CPU, FPGA, and GPU tracking demonstrators. We will also discuss why ATLAS ultimately selected a flexible CPU+GPU computing farm for Run 4, and its synergies with offline simulation.

Speaker: Punit Sharma (Brookhaven National Laboratory (US))