A fast and efficient SIMD track reconstruction algorithm for the LHCb Upgrade 1 VELO-PIX detector

TL;DR

This paper introduces a highly parallel SIMD-based track reconstruction algorithm for the LHCb Upgrade 1 VELO-PIX detector, capable of processing 4 TB/s data in real time, outperforming previous methods.

Contribution

The paper presents a novel SIMD-optimized parallel algorithm for vertex detector data reconstruction, improving speed and physics performance over existing scalar algorithms.

Findings

Significantly faster processing than previous algorithms

Enhanced physics performance in some cases

Effective utilization of SIMD instruction set extensions

Abstract

The upgraded CERN LHCb detector, due to start data taking in 2021, will have to reconstruct 4 TB/s of raw detector data in real time using commodity processors. This is one of the biggest real-time data processing challenges in any scientific domain. We present an intrinsically parallel reconstruction algorithm for the vertex detector of the LHCb experiment designed to optimally exploit multi-core general purpose architectures. We compare it to previous state-of-the-art scalar pattern recognition algorithms and show significantly faster processing and in some cases increased physics performance over all current alternatives. We evaluate the algorithm on two high-end architectures from two different vendors and discuss in detail the impact of different SIMD Instruction Set Architecture extensions on the performance.