Performance of a geometric deep learning pipeline for HL-LHC particle tracking
Abstract The Exa.TrkX project has applied geometric learning concepts such as metric learning and graph neural networks to HEP particle tracking. Exa.TrkX’s tracking pipeline groups detector measurements to form track candidates and filters them. The pipeline, originally developed using the TrackML...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2021-10-01
|
| Series: | European Physical Journal C: Particles and Fields |
| Online Access: | https://doi.org/10.1140/epjc/s10052-021-09675-8 |
| _version_ | 1818719123781189632 |
|---|---|
| author | Xiangyang Ju Daniel Murnane Paolo Calafiura Nicholas Choma Sean Conlon Steven Farrell Yaoyuan Xu Maria Spiropulu Jean-Roch Vlimant Adam Aurisano Jeremy Hewes Giuseppe Cerati Lindsey Gray Thomas Klijnsma Jim Kowalkowski Markus Atkinson Mark Neubauer Gage DeZoort Savannah Thais Aditi Chauhan Alex Schuy Shih-Chieh Hsu Alex Ballow Alina Lazar |
| author_facet | Xiangyang Ju Daniel Murnane Paolo Calafiura Nicholas Choma Sean Conlon Steven Farrell Yaoyuan Xu Maria Spiropulu Jean-Roch Vlimant Adam Aurisano Jeremy Hewes Giuseppe Cerati Lindsey Gray Thomas Klijnsma Jim Kowalkowski Markus Atkinson Mark Neubauer Gage DeZoort Savannah Thais Aditi Chauhan Alex Schuy Shih-Chieh Hsu Alex Ballow Alina Lazar |
| author_sort | Xiangyang Ju |
| collection | DOAJ |
| description | Abstract The Exa.TrkX project has applied geometric learning concepts such as metric learning and graph neural networks to HEP particle tracking. Exa.TrkX’s tracking pipeline groups detector measurements to form track candidates and filters them. The pipeline, originally developed using the TrackML dataset (a simulation of an LHC-inspired tracking detector), has been demonstrated on other detectors, including DUNE Liquid Argon TPC and CMS High-Granularity Calorimeter. This paper documents new developments needed to study the physics and computing performance of the Exa.TrkX pipeline on the full TrackML dataset, a first step towards validating the pipeline using ATLAS and CMS data. The pipeline achieves tracking efficiency and purity similar to production tracking algorithms. Crucially for future HEP applications, the pipeline benefits significantly from GPU acceleration, and its computational requirements scale close to linearly with the number of particles in the event. |
| first_indexed | 2024-12-17T20:01:57Z |
| format | Article |
| id | doaj.art-3869b806bd4f475fb2adf7edf5616d69 |
| institution | Directory Open Access Journal |
| issn | 1434-6044 1434-6052 |
| language | English |
| last_indexed | 2024-12-17T20:01:57Z |
| publishDate | 2021-10-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | European Physical Journal C: Particles and Fields |
| spelling | doaj.art-3869b806bd4f475fb2adf7edf5616d692022-12-21T21:34:27ZengSpringerOpenEuropean Physical Journal C: Particles and Fields1434-60441434-60522021-10-01811011410.1140/epjc/s10052-021-09675-8Performance of a geometric deep learning pipeline for HL-LHC particle trackingXiangyang Ju0Daniel Murnane1Paolo Calafiura2Nicholas Choma3Sean Conlon4Steven Farrell5Yaoyuan Xu6Maria Spiropulu7Jean-Roch Vlimant8Adam Aurisano9Jeremy Hewes10Giuseppe Cerati11Lindsey Gray12Thomas Klijnsma13Jim Kowalkowski14Markus Atkinson15Mark Neubauer16Gage DeZoort17Savannah Thais18Aditi Chauhan19Alex Schuy20Shih-Chieh Hsu21Alex Ballow22Alina Lazar23Lawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryLawrence Berkeley National LaboratoryCalifornia Institute of TechnologyCalifornia Institute of TechnologyUniversity of CincinnatiUniversity of CincinnatiFermi National Accelerator LaboratoryFermi National Accelerator LaboratoryFermi National Accelerator LaboratoryFermi National Accelerator LaboratoryUniversity of Illinois at Urbana-ChampaignUniversity of Illinois at Urbana-ChampaignPrinceton UniversityPrinceton UniversityUniversity of WashingtonUniversity of WashingtonUniversity of WashingtonYoungstown State UniversityYoungstown State UniversityAbstract The Exa.TrkX project has applied geometric learning concepts such as metric learning and graph neural networks to HEP particle tracking. Exa.TrkX’s tracking pipeline groups detector measurements to form track candidates and filters them. The pipeline, originally developed using the TrackML dataset (a simulation of an LHC-inspired tracking detector), has been demonstrated on other detectors, including DUNE Liquid Argon TPC and CMS High-Granularity Calorimeter. This paper documents new developments needed to study the physics and computing performance of the Exa.TrkX pipeline on the full TrackML dataset, a first step towards validating the pipeline using ATLAS and CMS data. The pipeline achieves tracking efficiency and purity similar to production tracking algorithms. Crucially for future HEP applications, the pipeline benefits significantly from GPU acceleration, and its computational requirements scale close to linearly with the number of particles in the event.https://doi.org/10.1140/epjc/s10052-021-09675-8 |
| spellingShingle | Xiangyang Ju Daniel Murnane Paolo Calafiura Nicholas Choma Sean Conlon Steven Farrell Yaoyuan Xu Maria Spiropulu Jean-Roch Vlimant Adam Aurisano Jeremy Hewes Giuseppe Cerati Lindsey Gray Thomas Klijnsma Jim Kowalkowski Markus Atkinson Mark Neubauer Gage DeZoort Savannah Thais Aditi Chauhan Alex Schuy Shih-Chieh Hsu Alex Ballow Alina Lazar Performance of a geometric deep learning pipeline for HL-LHC particle tracking European Physical Journal C: Particles and Fields |
| title | Performance of a geometric deep learning pipeline for HL-LHC particle tracking |
| title_full | Performance of a geometric deep learning pipeline for HL-LHC particle tracking |
| title_fullStr | Performance of a geometric deep learning pipeline for HL-LHC particle tracking |
| title_full_unstemmed | Performance of a geometric deep learning pipeline for HL-LHC particle tracking |
| title_short | Performance of a geometric deep learning pipeline for HL-LHC particle tracking |
| title_sort | performance of a geometric deep learning pipeline for hl lhc particle tracking |
| url | https://doi.org/10.1140/epjc/s10052-021-09675-8 |
| work_keys_str_mv | AT xiangyangju performanceofageometricdeeplearningpipelineforhllhcparticletracking AT danielmurnane performanceofageometricdeeplearningpipelineforhllhcparticletracking AT paolocalafiura performanceofageometricdeeplearningpipelineforhllhcparticletracking AT nicholaschoma performanceofageometricdeeplearningpipelineforhllhcparticletracking AT seanconlon performanceofageometricdeeplearningpipelineforhllhcparticletracking AT stevenfarrell performanceofageometricdeeplearningpipelineforhllhcparticletracking AT yaoyuanxu performanceofageometricdeeplearningpipelineforhllhcparticletracking AT mariaspiropulu performanceofageometricdeeplearningpipelineforhllhcparticletracking AT jeanrochvlimant performanceofageometricdeeplearningpipelineforhllhcparticletracking AT adamaurisano performanceofageometricdeeplearningpipelineforhllhcparticletracking AT jeremyhewes performanceofageometricdeeplearningpipelineforhllhcparticletracking AT giuseppecerati performanceofageometricdeeplearningpipelineforhllhcparticletracking AT lindseygray performanceofageometricdeeplearningpipelineforhllhcparticletracking AT thomasklijnsma performanceofageometricdeeplearningpipelineforhllhcparticletracking AT jimkowalkowski performanceofageometricdeeplearningpipelineforhllhcparticletracking AT markusatkinson performanceofageometricdeeplearningpipelineforhllhcparticletracking AT markneubauer performanceofageometricdeeplearningpipelineforhllhcparticletracking AT gagedezoort performanceofageometricdeeplearningpipelineforhllhcparticletracking AT savannahthais performanceofageometricdeeplearningpipelineforhllhcparticletracking AT aditichauhan performanceofageometricdeeplearningpipelineforhllhcparticletracking AT alexschuy performanceofageometricdeeplearningpipelineforhllhcparticletracking AT shihchiehhsu performanceofageometricdeeplearningpipelineforhllhcparticletracking AT alexballow performanceofageometricdeeplearningpipelineforhllhcparticletracking AT alinalazar performanceofageometricdeeplearningpipelineforhllhcparticletracking |