AI4Physics seminar: Nanosecond AI at the Large Hadron Collider

Abstract:

The Large Hadron Collider at CERN smashes protons together at nearly the speed of light to probe the fundamental building blocks of the universe. Each collision produces a shower of particles — and a flood of data. In fact, the LHC generates tens of thousands of exabytes of raw data every year, far more than can be stored or analyzed directly.

To make discoveries like the Higgs boson possible, scientists rely on advanced AI and machine learning systems, combined with specialized hardware, to sift through millions of events per second in real time. These tools help identify rare and valuable signals hidden in a sea of noise.

In this talk, we will discuss how physicists and engineers use AI to push the limits of data, speed, and precision in one of the most ambitious scientific experiments ever undertaken.

Bio:

Thea Aarrestad is a fellow at the Institute for Particle Physics and Astrophysics at ETH Zürich. She holds a PhD in Particle Physics from the University of Zürich and has worked as a research fellow at CERN in Geneva before moving to ETH. Her research centers on how Machine Learning can be applied to particle physics problems, especially focusing on using real-time Machine Learning (ML) for discovering new physics phenonema. She has worked on tools for performing low-power, nanosecond ML inference on field-programmable gate arrays (FPGAs), as well as developing new methods for collecting and analyzing proton collision data at the CERN Large Hadron Collider using ML-based anomaly detection techniques.