CE31 - Physique subatomique et astrophysique

Artificial intelligence on FPGAs: a breakthrough for data acquisition in high energy physics experiments and beyond – AIDAQ

Submission summary

The growing processing power of high-end FPGAs (Field Programmable Gate Array) opens new cutting-edge technological opportunities combining the powerful potential of Artificial Intelligence algorithms to big data processing capabilities (100Tb/s). Overcoming this major technological challenge opens new paths in data processing in particle physics experiments with high potential for general purpose applications. At CERN, the upgrade of the Large Hadron Collider and of its detectors represents a major milestone for both fundamental physics and advanced technological developments. The increased luminosity of the LHC requires an upgrade of the ATLAS detector. This upgrade is a cornerstone of the European strategies for High Energy Physics. Upgrading the readout electronics of the ATLAS liquid Argon Calorimeter is an essential step which necessitates state-of-the-art FPGA boards to instantly process the huge volume of data with sophisticated algorithms which are needed to maintain the accuracy required for physics analyses at high luminosity. The ATLAS group at CPPM is responsible of developing those readout boards based on a long-term experience with high-end FPGAs. This group has an extensive experience with machine learning techniques and artificial intelligence algorithms applied to physics analyses.  This project is a unique opportunity for a breakthrough in data processing for High Energy Physics experiments and opens up a path with great potential for a wide range of applications.
With the increased luminosity of the LHC, the number of multiple proton-proton collisions in one bunch crossing (called pileup) increases significantly putting more stringent requirements on the LHC detectors electronics and data processing. The physics performance of the currently used algorithms degrade in high pileup conditions, namely the resolution on the energy deposited in the calorimeter. An excellent resolution on the deposited energy and an accurate detection of the deposited time, in the blurred environment created by the pileup, is crucial for the operation of the calorimeters and of the full ATLAS detector to enhance its discovery potential. Artificial intelligence algorithms have proven to be very powerful tools in data processing and provide the most interesting candidate to recover the performance of the current common algorithms in high pileup conditions. The main purpose of this project is to implement artificial intelligence and machine learning techniques to dramatically enhance the physics performance of the liquid Argon calorimeter data processing in high luminosity conditions. One of the main challenges is to efficiently implement these techniques into the dedicated data acquisition electronics, based on high end FPGAs. Then, to enlarge the scope of the project, algorithms used in general applications, such as civil security, survey drones, autonomous cars..., will be deployed, tested and validated using the same software tools and hardware to open up the path towards cutting edge industrial applications.

Project coordination

Georges Aad (Centre National de la Recherche Scientifique Délégation Provence et Corse_Centre de physique des particules de Marseille)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.


CNRS DR12_CPPM Centre National de la Recherche Scientifique Délégation Provence et Corse_Centre de physique des particules de Marseille

Help of the ANR 250,344 euros
Beginning and duration of the scientific project: - 36 Months

Useful links

Explorez notre base de projets financés



ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter