Development and construction of an FPGA based track finder for the CMS Level 1 trigger.
Lead Research Organisation:
Brunel University London
Department Name: Electronic and Computer Engineering
Abstract
This project is to help develop and construct an FPGA based track finding system for use in the Level 1 trigger system of the CMS detector at the CERN-LHC. This is part of the suite of upgrades planned for the CMS detector in order to prepare it for high luminosity running, which is expected to take place after Long Shutdown 3 (that is, around 2024). This will represent the first time CMS has used tracking information at Level 1, which will be essential for managing data rates post upgrade.
The focus on this project will be on developing the firmware required for the system and, in particular, that required for the Track Finder boards, whose role is to perform track reconstruction and fitting, using data from specially designed layers of the upgraded Tracker, which will be installed within CMS at the same time.
The focus on this project will be on developing the firmware required for the system and, in particular, that required for the Track Finder boards, whose role is to perform track reconstruction and fitting, using data from specially designed layers of the upgraded Tracker, which will be installed within CMS at the same time.
People |
ORCID iD |
| Joanne Elise Cole (Primary Supervisor) | |
| Maziar Ghorbani (Student) |
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| ST/N503964/1 | 01/10/2015 | 31/03/2021 | |||
| 2010696 | Studentship | ST/N503964/1 | 01/01/2018 | 30/06/2021 | Maziar Ghorbani |
| Title | High-Level Synthesis - an automated design process for HEP expriments |
| Description | High-Level Synthesis is an automated design process that interprets an algorithmic description of the desired behaviour and creates digital hardware that implements that behaviour. The tool has been used in creating hardware for FPGA based project since 1994, however, recently has been deployed in CMS experiments and High-Luminosity LHC upgrades of trigger algorithm. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2016 |
| Provided To Others? | Yes |
| Impact | The tool is used for benchmarking against the VHDL or Verilog implementations of the current algorithm against the accuracy, latency, resource usage, and code size. With the use of the tool, physicists are able to gather estimates for firmware implementations without using hardware description languages and accelerating the design process of trigger firmware implementation in terms of consistency of its corresponding C++ model. |
| Title | High-Level Synthesis of Linear Regression algorithm in CMS Track-Finder expriments |
| Description | The CMS detector at LHC experiment generates data at a very high rate. Not all of this data contains interesting physics to be stored for further analysis. A fast data processing model is required to filter the data and reduced the rate. The track-fitter and track-finder chain are responsible to analyse and eliminate a large amount of data according to the specifications for on-line processing. The chain is constructed from several sequential modules. Every module is designed to identify the interesting and pass the data to the next module. The linear regression module along the chain finds and filters the particles that are furthest away from the tracks fitted by linear regression algorithm. The design must be fully pipelined to comply with the data rate and latency requirements and specification by the system. The synthesis tool allows fast implementation of complex physics algorithm and testing of resources usage and timing constraints in firmware design. |
| Type Of Material | Data handling & control |
| Year Produced | 2016 |
| Provided To Others? | Yes |
| Impact | The impact of using the model is examining the feasibility of using high-level synthesis in high energy physics experiments and future projects. |
| Description | Phase 2 Tracker Lelvel 1 Track-Finder Algorithm (SW+FW) |
| Organisation | European Organization for Nuclear Research (CERN) |
| Country | Switzerland |
| Sector | Academic/University |
| PI Contribution | The collaboration is between several organisations in connection to the CMS Inner and Outer Tracker Backend Data Processing Systems project, covering the development of the off-detector systems for readout and control of the tracker, including Level 1 track finding and reconstruction. The Phase 2 Tracker L1 TF Algo (SW+FW) is a sub-group of the Tracker BE Data Processing Systems (DPS) WG to define and develop the reference L1 track finding algorithm for the Tracker BE specification. The collaboration holds several meetings, talks and conferences where the team members can participate and share their research outcome. I have been attending the events and taking part in available tasks, specifically, the development of alternative track-finder linear regression module using High-Level Synthesis (HLS) and automated design process that interprets the current algorithm and creates digital hardware for FPGA based firmware design. |
| Collaborator Contribution | The contributions include: Define and development of Level 1 Track Finding Algorithm for CMS Tracker, Level 1 track-finder HLS algorithm implementation and optimisations, Development of hybrid software and hybrid HLS firmware, Creating task-groups and organising meetings, run tests and simulations to ensure correct functionality of the designs. |
| Impact | Development and optimisation of an FPGA based track finder module using the linear regression model and high-level synthesis algorithm. (in progress) |
| Start Year | 2017 |
| Description | Phase 2 Tracker Lelvel 1 Track-Finder Algorithm (SW+FW) |
| Organisation | Rutherford Appleton Laboratory |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The collaboration is between several organisations in connection to the CMS Inner and Outer Tracker Backend Data Processing Systems project, covering the development of the off-detector systems for readout and control of the tracker, including Level 1 track finding and reconstruction. The Phase 2 Tracker L1 TF Algo (SW+FW) is a sub-group of the Tracker BE Data Processing Systems (DPS) WG to define and develop the reference L1 track finding algorithm for the Tracker BE specification. The collaboration holds several meetings, talks and conferences where the team members can participate and share their research outcome. I have been attending the events and taking part in available tasks, specifically, the development of alternative track-finder linear regression module using High-Level Synthesis (HLS) and automated design process that interprets the current algorithm and creates digital hardware for FPGA based firmware design. |
| Collaborator Contribution | The contributions include: Define and development of Level 1 Track Finding Algorithm for CMS Tracker, Level 1 track-finder HLS algorithm implementation and optimisations, Development of hybrid software and hybrid HLS firmware, Creating task-groups and organising meetings, run tests and simulations to ensure correct functionality of the designs. |
| Impact | Development and optimisation of an FPGA based track finder module using the linear regression model and high-level synthesis algorithm. (in progress) |
| Start Year | 2017 |
| Title | Linear Regression module for Phase 2 CMS Tracker Level 1 Track-Finder |
| Description | The linear regression model implementation in the CMS tracker is well studied and tested for off-line data processing and analysis, however, just recently tried and tested for online track fitting and track reconstruction suitable for firmware and System on Chip (SoC) design. |
| Type Of Technology | New/Improved Technique/Technology |
| Year Produced | 2019 |
| Impact | The software implements and tests existed linear regression algorithm in CMSSW environment to High-Level Synthesis (HLS) version of the algorithm compatible with several FPGA chips. The generated Intellectual Property (IP) can replace the offline version of the linear regression module in the track-finder chain for testing and simulations of the online version while the data are generated from the CMS detector. |
| Description | ISOTDAQ 2019 - International School of Trigger and Data AcQuisition |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Study participants or study members |
| Results and Impact | participants to high energy physics, engineering and computing of triggering and acquiring data for physics experiments. The main aim of the school is to provide an overview of the basic instruments and methodologies used in particle physics, ranging from small experiments in the lab to the very large LHC experiments with different levels of complexity. The impact of attending the event was broadening the knowledge of instrumentation in high energy physics and creating a large circle of connections from researchers with similar interests to professional experts in the field that are essential for current experiments and future research. |
| Year(s) Of Engagement Activity | 2019 |
| URL | https://indico.cern.ch/event/828931 |
| Description | Long term attachment at the European Organization for Nuclear Research (CERN) |
| Form Of Engagement Activity | Participation in an open day or visit at my research institution |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Study participants or study members |
| Results and Impact | A long term attachment is a single visit of 365 days at the European Organization for Nuclear Research (CERN). It provides a great opportunity for PhD researchers to work in a professional environment, participate in experimental physics tasks and engage in activities such as conferences, seminars and talks. The impact of the visit is broadening the student's knowledge and observing how large projects are managed and implemented. |
| Year(s) Of Engagement Activity | 2019,2020 |