在LHCb的GPU驻留HLT1中部署混合PVFinder算法
Deploying a Hybrid PVFinder Algorithm for Primary Vertex Reconstruction in LHCb's GPU-Resident HLT1
AI总结：
本文提出在LHCb GPU驻留HLT1中部署混合PVFinder算法，通过优化神经网络处理提升顶点重建效率。
摘要：
LHCb的Run 3升级引入了一个完全基于软件的触发系统，以30MHz的速度运行，每束束流交叉处理平均5.6个质子-质子碰撞顶点（事件）。本文介绍了为PVFinder开发的推理引擎，PVFinder是一种混合深度神经网络，用于寻找初级顶点，这些顶点是所有后续粒子衰变的起源。该整合解决了关键的实时约束，包括固定内存池、单流执行以及NVIDIA GPU上每个事件处理预算不超过400微秒。我们介绍了一层翻译层，将Allen的数组结构（SoA）数据布局与cuDNN的张量格式连接起来，同时保持零拷贝语义和确定性行为。当前性能显示CNN阶段贡献了显著的吞吐量开销。我们提出了一个路线图，旨在通过混合精度计算、模型压缩和其他技术实现数量级的改进。
英文摘要：
LHCb's Run 3 upgrade introduced a fully software-based trigger system operating at 30~MHz, processing an average of 5.6 proton-proton collision vertices per bunch crossing (event). This work presents the development of an inference engine for PVFinder, a hybrid deep neural network for finding primary vertices, the proton-proton collision points from which all subsequent particle decays originate into Allen, LHCb's High Level Trigger (HLT1) framework. The integration addresses critical real-time constraints including fixed memory pools, single-stream execution, and sub-400~$μ$s per-event processing budgets on NVIDIA GPUs. We introduce a translation layer that bridges Allen's Structure-of-Arrays (SoA) data layout with cuDNN's tensor format while maintaining zero-copy semantics and deterministic behavior. Current performance shows the CNN stage contributes significant throughput overhead. We present a roadmap targeting order-of-magnitude improvements through mixed-precision computing, model compression and other techniques.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
