Jump to: Fellows - Projects - Presentations - Publications - Videos

Innovative Algorithms

Algorithms to perform the real-time processing in the trigger and the reconstruction of both real and simulated detector data are critical components of HEP’s computing challenge. These algorithms face a number of new challenges in the next decade due to new and upgraded accelerator facilities, detector upgrades and new detector technologies, increases in anticipated event rates, and emerging computing architectures. Projects in the IA area are focused in three areas:

  • Developing tracking algorithms for the HL-LHC
  • Re-engineering algorithms for hardware accelerators
  • Exploiting major advances in machine learning

Tracking for the HL-LHC is an area in particular need of novel approaches as it is expected to require a large fraction of the CPU budget. IA is working to develop more efficient and performant tracking algorithms. Hardware accelerators are expected to be the way forward to speed up reduce infrastructure cost. IA is exploring the use of hardware accelerators for tracking and the use of ML on accelerators in realistic HEP applications. In the area of machine learning, IA aims to capitalize on industry and data science techniques and tools. In particular, we are investigating new HEP applications of ML and to apply new ML techniques to HEP.

Contact us: ia-team@iris-hep.org

Card image capMarkus Atkinson
Card image capGowtham Atluri
Card image capThomas Boettcher
Card image capDaniel Craik
Card image capIrina Espejo
Card image capLouis-Guillaume Gagnon
Card image capElyssa Hofgard
Card image capAniket Khanal
Card image capSlava Krutelyov
Card image capDylan Rankin
Card image capKate Richardson
Card image capGarima Singh
Card image capMarian Stahl
Card image capMatevz Tadel
Card image capCarlo Varni
Card image capBeomki Yeo

Current and Previous IA Fellows

IA Projects


Card image cap

Accelerated GNN Tracking

accel-gnn-tracking
More information
Exploratory
Card image cap

Accelerators and ML for reconstruction

Accelerated calorimeter reconstruction using Machine Learning as a Service
More information
Archived
Card image cap

ACTS

Development of experiment-independent, thread-safe track reconstruction.
More information
Development
Card image cap

exploratory-ml

Analysis Reinterpretation
More information
Development
Card image cap

GPU Trigger Project

Allen: a GPU trigger for LHCb
More information
Testing
Card image cap

Line-Segment tracking

Segment linking tracking for CMS
More information
Development
Card image cap

Machine Learning for jets

Machine learning for jets
More information
Development
Card image cap

mkFit

Modernizing Kalman filter tracking for CMS
More information
Deployed
Card image cap

PV-Finder

CNNs to find primary vertices
More information
Testing

IA Presentations

IA Publications

Recent IA Recordings

Several on IRIS-HEP and HL-LHC tracking

3 May 2022
Several on IRIS-HEP and LHCb

3 May 2022
Sofia Graziano on Rotationally Equivariant Graph Neural Networks

1 Nov 2021
Kyle Feist on Evolutionary Algorithm for Optimization of Track Reconstruction at a Muon Collider

1 Nov 2021
Eric Moreno on Anomaly Detection with Spiking Neural Networks

27 Oct 2021
Rohith Karur on ANNs for ACTS

18 Oct 2021

Join us

We collaborate with groups around the world on code, data, and more. See our project pages for more.

gantt title Innovative Algorithms project lifecyles dateFormat YYYY-MM-DD axisFormat %Y Accelerated GNN Tracking :done , 2019-09-12, 2022-12-04 Accelerators and ML for reconstruction :crit , 2019-01-01, 2021-10-01 ACTS :done , 2016-06-09, 2022-12-04 GPU Trigger Project :done , 2018-04-19, 2022-12-04 Line-Segment tracking :done , 2020-09-01, 2022-12-04 mkFit :active , 2014-02-23, 2022-12-04 PV-Finder :done , 2018-01-19, 2022-12-04