5D calorimetry at the HL-LHC: hard real-time embedded architectures for system testing and production
PT-CERN Call 2022/2
Instituto Superior Técnico (Universidade de Lisboa)
CERN Experimental Physics Department
The upcoming High-Luminosity phase of the Large Hadron Collider (HL-LHC) requires the replacement of the CMS detector endcap calorimeters with the new High-Granularity Calorimeter (HGCAL). The HGCAL detector will comprise more than 6 million channels to provide precise sensing and measurement of 3D position, timing, and energy of particles produced in the collisions. The HGCAL electronics comprise a large and rather complex processing system, whose capability will have to be scaled by several orders of magnitude to satisfy the demands required by this upgrade and lay out the blueprint for future detectors. In particular, about 100 Field Programmable Gate Arrays (FPGAs) will be used in the off-detector back-end to perform the configuration, timing, control, and monitoring of the on-detector ASIC electronics. While several IP blocks have been developed, they only provide proof-of-concept functionality for the different functions and are just a first step on the way to the full 100-FPGA detector system. Those blocks need to be tested and integrated into increasingly larger systems. A rather demanding challenge that is still to be tackled is the scaling up of these IPs, from their existing (small scale) prototypes. Integrating them in gradually larger and more complex systems as needed to support the construction of the detector until 2028 is not a trivial task, as reliability needs to be ensured at every single step. Under these premises, the proposed PhD workplan focuses on the definition of architectures and attending testing frameworks to support the scalability and the integration of the several (existing and/or to be defined) HGCAL data-processing elements and control systems. The devised (modular) architectures will have to comply with the demanding requisites (in terms of data throughput and interface) that characterize the heterogeneous nature of the HGCAL data processing ecosystem.