BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Machine Learning for Trigger and Data Acquisition
DTSTART:20260722T113000Z
DTEND:20260722T123000Z
DTSTAMP:20260916T103700Z
UID:indico-event-1690265@indico.cern.ch
DESCRIPTION:Speakers: Thomas Owen James (CERN)\n\nAbstract\nThe LHC produc
 es billions of collisions per second during operation\, equivalent to Pb/s
  of raw data for the largest experiments CMS and ATLAS. It would be imposs
 ible to readout\, process or store all of this data. A multi-tier trigger 
 system is therefore used to select only a tiny fraction of the most intere
 sting collisions for further analysis\, with a high efficiency and low fal
 se positive rate. As a result of the throughput and latency constraints\, 
 the first level of this trigger must typically make a selection decision w
 ithin a few microseconds. In the meantime\, Machine learning (ML) is a ric
 h and exciting field of research\, constantly inventing new and more power
 ful techniques. At the same time\, developers are supporting the growth of
  ML with faster\, more parallel processors and devices designed specifical
 ly for ML. In light of this\, deploying ML into the real-time processing f
 or trigger and data acquisition  is becoming increasingly possible and r
 elevant. As the LHC upgrades by around a factor of 5 in instantaneous lumi
 nosity for the next decade\, this ‘fast ML’ at the edge will be undoub
 tedly required to reduce and filter the vast amounts of data. This lecture
  covers a recap of neural networks (NN) and the tools and frameworks used 
 to implement them\, before diving into a multitude of examples where ML is
  already being used and/or developed for ultra low-latency event selection
 \, fast reconstruction\, anomaly detection\, and data reduction/filtering\
 , at the LHC experiments. The implementation of “real-time” machine le
 arning inference on GPU and FPGA devices will be explored\, and the latest
  tools and tricks for optimisation in the space\, such as high level synth
 esis\, quantisation\, and knowledge distillation will be discussed. \nBio
 \n\nThomas is an applied physicist in the CMS group of CERN\, where he app
 lies machine learning (ML) solutions to ultra low-latency (microsecond) da
 ta-processing using field programmable gate arrays (FPGAs). Thomas has wor
 ked on the CMS experiment since 2013\, specialising in data acquisition\, 
 fast particle selection\, and event reconstruction in FPGAs. He is the CER
 N Openlab CTO for AI and Edge devices\, and has worked on multiple Openlab
  projects with Micron Technology on real-time event selection using deep l
 earning accelerators\, and Compute Express Link (CXL)-based shared memory.
 \nThomas joined the CERN EP-CMD group in 2019. In 2018\, Thomas graduated 
 with his PhD in particle physics from Imperial College London\, UK\, where
  he developed a novel FPGA-based online particle track finder for CMS at t
 he High Luminosity LHC\, earning him the CMS Thesis Award for that year. P
 rior to that\, he obtained his Masters and Bachelors degree in physics wit
 h theoretical physics from the same institute.\nPlease note that pictures 
 and videos might be taken during the event. The pictures and videos might 
 be used for communication about the event. By joining the lecture\, you ar
 e agreeing to being featured in these communication actions. \n\n\nhttps:
 //indico.cern.ch/event/1690265/
LOCATION:31/3-004 - IT Amphitheatre (CERN)
URL:https://indico.cern.ch/event/1690265/
END:VEVENT
END:VCALENDAR