The DPG Spring Meeting in Bonn had to be cancelled! Read more ...
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 46: Detector systems III
T 46.9: Talk
Wednesday, April 1, 2020, 18:35–18:50, H-HS II
Decoding of SciFi detector raw data in the Allen GPU-based High Level Trigger for the LHCb Upgrade experiment — •Lars Funke and Holger Stevens — Experimentelle Physik 5, TU Dortmund
In the coming LHC run, the LHCb experiment will take data without a hardware trigger stage, increasing the input rate to the software-based High Level Trigger (HLT) by a factor of 30. Additionally, the number of pp interactions per bunch crossing will increase fivefold. Both of those facts pose a challenge to the HLT, as the projected throughput implies the need for a large amount of computing power.
The Allen project implements the first stage of the HLT on Graphics Processing Units (GPUs), which offer a considerable performance-per-price advantage over CPUs, if used with suitable algorithms. A significant time share of the trigger sequence is used by raw data decoding, as for some of the subdetectors, non-trivial transformations are necessary.
In this talk, the algorithms and challenges of the Scintillating Fibre Tracker (SciFi) raw data decoding are outlined.