A graph neural network (ETX4VELO) reconstructs LHCb VELO tracks with performance comparable to the production 'search by triplet' algorithm while running end to end in the GPU-based first-level trigger, with additional FPGA power and throughput comparisons.
Computing and software for LHCb Upgrade II
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
A second major upgrade of the LHCb experiment is necessary to allow full exploitation of the High Luminosity LHC for flavour physics. The new experiment will operate in Run 5 of the LHC at a luminosity up to $1.5\times 10^{34}cm^{-2}s^{-1}$. The experiment will therefore experience extremely high particle fluences and data rates, posing a high challenge not only for the detector but also for the software and computing resources needed to readout, reconstruct, select and analyse the data. This document presents these challenges and the ongoing and future R&D programme necessary to address them. This programme will benefit not only the LHCb Upgrade II experiment, but the whole particle physics community as similar challenges will be faced by the next generation of experiments.
fields
physics.data-an 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Real-Time Analysis of Unstructured Data with Machine Learning on Heterogeneous Architectures
A graph neural network (ETX4VELO) reconstructs LHCb VELO tracks with performance comparable to the production 'search by triplet' algorithm while running end to end in the GPU-based first-level trigger, with additional FPGA power and throughput comparisons.