IQM Seminar
IQM讲座

[2026年5月26日] TOU DA YU——Allen: LHCb’s GPU-based High Throughput Software Trigger

2026-05-25 11:51:00 来源:华南师范大学量子物质研究院 点击: 收藏本文

报告题目:Allen: LHCb’s GPU-based High Throughput Software Trigger


报告人:Dr. TOU DA YUTsinghua University


报告时间:2026526日 上午10:00-11:00


报告地点:理8栋118学术报告厅


摘要:

Allen, LHCb’s GPU-based software trigger, operates LHCb’s High Level Trigger 1 (HLT1) system. It processes the LHCb’s full detector readout at 30 MHz, at a data rate of 4 Tb/s, on a computing farm of about 500 GPUs. This makes LHCb one of the world’s first high energy physics experiment to operate without a hardware trigger. Today, I will present Allen, discussing how the framework was designed to meet the demanding physics and computing performance requirements. I will also cover LHCb’s operational experience in successfully taking data with this trigger, as well as our future plans.


报告人简介:

Da Yu Tou received his PhD from Sorbonne University in 2020, focusing on test of lepton universality in b->sll decays at the LHCb experiment. Now a postdoctoral research at Tsinghua University, his recent research has centered around hadron spectroscopy via amplitude analysis and LHCb’s Real-Time Analysis (RTA) software trigger, specializing in its GPU-based High Level Trigger 1 (HLT1). He currently serves as a convener for the LHCb B-hadrons and quarkonia physics working group and is a convener for the HEP Software Foundation (HSF) in the Reconstruction and Software Triggers activity area.