• DocumentCode
    1096669
  • Title

    A parallel pipelined dataflow trigger processor

  • Author

    Lee, C. ; Miller, G. ; Kaplan, D.M. ; Sa, J. ; Hsiung, Y.B. ; Carey, T. ; Jeppesen, R.

  • Author_Institution
    Northern Illinois Univ., DeKalb, IL, USA
  • Volume
    38
  • Issue
    2
  • fYear
    1991
  • fDate
    4/1/1991 12:00:00 AM
  • Firstpage
    461
  • Lastpage
    470
  • Abstract
    A parallel pipelined dataflow trigger processor which is used in Fermilab E789 is described. E789 is an experiment designed to study low-multiplicity decays of particles containing b or c quarks. The processor consists of an upstream vertex processor and a downstream track processor. The algorithms which reconstruct the postulated particle paths and calculate particle origin are implemented via interconnected function-specific hardware modules. The algorithm is directly dependent upon the organization of the modules, the specific arrangement of the intermodule cabling, the on-board wire patch patterns, and, where appropriate, the on-board memory data. The uniqueness of this particular dataflow processor lies in the inclusion of the particle origin calculation. The vertex processor utilizes silicon detector hits to calculate the particle origin and to verify consistency with the downstream track. The downstream tracks are reconstructed by the track processor. The processor provides an indication of the presence of at least one interesting particle pair in the current event by asserting read on its read/skip output. The read assertion is then used as a trigger to capture all of the event´s data for subsequent extensive offline analysis
  • Keywords
    parallel processing; physics computing; E789; algorithm; downstream track processor; parallel pipelined dataflow trigger processor; upstream vertex processor; Detectors; Hardware; Laboratories; Particle tracking; Physics; Protocols; Silicon; Spectroscopy; Target tracking; Wire;
  • fLanguage
    English
  • Journal_Title
    Nuclear Science, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9499
  • Type

    jour

  • DOI
    10.1109/23.289340
  • Filename
    289340