Speaker
Description
Accurate alignment of detector elements in real-time is essential to maintain the integrity of reconstructed particle trajectories, especially in high-rate environments like the ATLAS experiment at the Large Hadron Collider (LHC). Any misalignment in the detector geometry can introduce systematic biases and potentially affect the accuracy of precision physics measurements. Current calibration systems that correct for these effects require substantial computational resources and these methods often lead to high operational costs and are often unable to handle rapidly changing conditions, leading to systematic inaccuracies and potential biases in physics measurements.
To overcome these challenges, we propose a calibration system that employs a lightweight neural network to predict the misalignment of the detectors in real time. We present a neural network model with hierarchical subset solvers optimized for Versal ACAP that predicts detector misalignment based on the detector’s current geometry and statistical characteristics of reconstructed particle tracks.
We address this by implementing the system on a heterogeneous architecture, partitioning sequential tasks to CPUs while offloading the computationally intensive matrix multiplications to the AI Engines to exploit their specialized vector processing capabilities. We benchmark this implementation against FPGA and GPU to evaluate trade-offs in latency and power across different data types. This work is implemented as a proof of concept for a modern beam telescope to demonstrate the viability of performing inference on the edge at low cost per watt- a crucial requirement for future high-energy physics experiments.
| Minioral | Yes |
|---|---|
| IEEE Member | Yes |
| Are you a student? | No |