The Muon to Central Trigger Processor Interface for the Upgrade of the ATLAS Muon Trigger for Run-3

19 Sept 2018, 12:00
25m
CAR 1.09 (aula)

CAR 1.09 (aula)

Oral Trigger Trigger

Speaker

Marcos Vinicius Silva Oliveira (Federal University of Juiz de Fora (BR))

Description

The Muon-to-Central Trigger Processor Interface(MUCTPI) of the Level-1 muon trigger of the ATLAS experiment is being replaced for the LHC Run-3. The upgraded MUCTPI is implemented as an ATCA module using high-end FPGAs and high-density ribbon fibre-optic modules to integrate over 270 multi-gigabit optical inputs and outputs on a single board.
The MUCTPI also features a System-on-Chip(SoC) with an ARM processor running an embedded Linux OS to control the module. We present results from the hardware and firmware validation of the second prototype based on Xilinx/Ultrascale+ FPGAs, and about the SoC used to interface to the experiment run control system.

Summary

The Muon-to-Central Trigger Processor Interface (MUCTPI) is part of
the Level-1 trigger system of the ATLAS experiment at CERN. It
receives and combines information on muon candidates from the 208
trigger sector logic modules of the detector and calculates the
candidate multiplicity, taking into account the possible double
counting between trigger sectors due to the geometrical overlap of the
muon chambers and the trajectory of the muons in the magnetic field,
and sends the result to the Central Trigger Processor (CTP). The
existing MUCTPI is being upgraded for Run-3 of the Large Hadron
Collider, in order to interface with the new muon endcap trigger
sector logic modules which will be deployed as part of the muon new
small wheel upgrade. The upgraded MUCTPI will also be able to send
full precision information on the muon candidates identified by the
muon trigger processors at the bunch crossing rate to the topological
trigger processor (L1Topo) of the Level-1 trigger system, which will
allow combined calorimeter/muon topological trigger algorithms to be
implemented. The MUCTPI for Run-3 is implemented as a single
AdvancedTCA (ATCA) blade, replacing a full 9U VME shelf with 18 boards
of the existing system. This high level of integration is enabled by
the use of state-of-the-art FPGA devices, featuring a large number of
on-chip high-speed serial transceivers, and high-density ribbon
fibre-optics receiver and transmitter modules. The module features
over 270 multi-gigabit optical inputs/outputs operating at line rates
between 6.4 and 11.2 Gb/s, resulting in an aggregate bandwidth of 2
Tb/s. Two large Xilinx Virtex UltraScale(+) FPGAs (sector processors),
each covering one side of the detector, are used to receive and
process the muon trigger information from the 208 sector logic modules
and forward lists of muon candidates to L1Topo through up to 48 serial
optical links. A third FPGA, a Xilinx Kintex UltraScale device, is
used to merge the information from the two sides of the detector, to
perform the required trigger calculations and to send the resulting
object multiplicities and trigger flags to the CTP. For events
accepted by the CTP, it also outputs a list of muon candidates to the
DAQ and high-level trigger systems. Finally, a Xilinx Zynq
System-on-Chip (SoC) FPGA is used to interface the MUCTPI to the ATLAS
run control system through a Gigabit Ethernet connection. It is used
to configure and control the MUCTPI and to read state and monitoring
information. The SoC device runs an embedded Linux with
application-specific software using a remote-procedure-call approach.
In addition, we have ported a simple run control application from the
ATLAS TDAQ software which can control and configure the MUCTPI,
provide monitoring data (including ROOT histograms). A first fully
functional prototype of the MUCTPI exists and has been extensively
tested. We present results from the validation of second prototype,
which uses 16 nm Ultrascale+ FPGAs for the sector processors. We also
give an overview of the firmware as well as the software environment
running on the SoC and show results from first integration tests.

Primary author

Marcos Vinicius Silva Oliveira (Federal University of Juiz de Fora (BR))

Presentation materials