GPU real-time processing in NA62 trigger system

10th International Conference on Kaon Physics

Ammendola, R.^1 ; Biagioni, A.^2 ; Chiozzi, S.^3 ; Cretaro, P.^2 ; Di Lorenzo, S.^4,5 ; Fantechi, R.^5 ; Fiorini, M.^3,6 ; Frezza, O.^2 ; Lamanna, G.^7 ; Lo Cicero, F.^2 ; Lonardo, A.^2 ; Martinelli, M.^2 ; Neri, I.^3 ; Paolucci, P.S.^2 ; Pastorelli, E.^2 ; Piandani, R.^5 ; Piccini, M.^8 ; Pontisso, L.^5 ; Rossetti, D.^9 ; Simula, F.^2 ; Sozzi, M.^4,5 ; Vicini, P.^2

INFN, Sezione di Roma Tor Vergata, Roma, Italy^1

INFN, Sezione di Roma, Roma, Italy^2

INFN, Sezione di Ferrara, Ferrara, Italy^3

Universita Degli Studi di Pisa, Pisa, Italy^4

INFN, Sezione di Pisa, Pisa, Italy^5

Università Degli Studi di Ferrara, Ferrara, Italy^6

INFN LNF, Frascati, Italy^7

INFN, Sezione di Perugia, Perugia, Italy^8

NVIDIA Corp., Santa Clara

CA, United States^9

关键词: Commercial graphics;    MultiRing;    Parallel com- puting;    Real time;    Realtime processing;    Ring Imaging Cherenkov detector;    Trigger and data acquisition system;    Trigger systems;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/800/1/012046/pdf
DOI  :  10.1088/1742-6596/800/1/012046

来源: IOP

【 摘 要 】

A commercial Graphics Processing Unit (GPU) is used to build a fast Level 0 (L0) trigger system tested parasitically with the TDAQ (Trigger and Data Acquisition systems) of the NA62 experiment at CERN. In particular, the parallel computing power of the GPU is exploited to perform real-time fitting in the Ring Imaging CHerenkov (RICH) detector. Direct GPU communication using a FPGA-based board has been used to reduce the data transmission latency. The performance of the system for multi-ring reconstrunction obtained during the NA62 physics run will be presented.