Within the Phase-II upgrade of the LHC, the readout electronics of the ATLAS LAr Calorimeters
are prepared for high luminosity operation expecting a pile-up of up to 200 simultaneous proton-
proton interactions. Moreover, the calorimeter signals of up to 25 subsequent collisions are
overlapping, which increases the difficulty of energy reconstruction. Real-time processing of
digitized pulses sampled at 40 MHz is performed using FPGAs.
To cope with the signal pile-up, new machine learning approaches are explored: convolutional and
recurrent neural networks outperform the optimal signal filter currently used, both in assignment
of the reconstructed energy to the correct bunch crossing and in energy resolution.
Very good agreement between neural network implementations in FPGA and software based
calculations is observed. The FPGA resource usage, the latency and the operation frequency are
analyzed. Latest performance results and experience with prototype implementations are analyzed
and are found to fit the requirements for the Phase-II upgrade.