Abstract
Real-time data processing is one of the central processes of particle physics experiments which require large computing resources. The LHCb (Large Hadron Collider beauty) experiment will be upgraded to cope with a particle bunch collision rate of 30 million times per second, producing $10^9$ particles/s. 40 Tbits/s need to be processed in real-time to make filtering decisions to store data. This poses a computing challenge that requires exploration of modern hardware and software solutions. We present Compass, a particle tracking algorithm and a parallel raw input decoding optimised for GPUs. It is designed for highly parallel architectures, data-oriented and optimised for fast and localised data access. Our algorithm is configurable, and we explore the trade-off in computing and physics performance of various configurations. A CPU implementation that delivers the same physics performance as our GPU implementation is presented. We discuss the achieved physics performance and validate it with Monte Carlo simulated data. We show a computing performance analysis comparing consumer and server grade GPUs, and a CPU. We show the feasibility of using a full GPU decoding and particle tracking algorithm for high-throughput particle trajectories reconstruction, where our algorithm improves the throughput up to 7.4$\times$ compared to the LHCb baseline.
Highlights
High-energy physics experiments produce large data streams that must be processed, filtered, and analyzed
This paper presents the implementation of a dataoriented approach, focusing on creating algorithms for SIMD (Single Instruction Multiple Data) architectures, minimizing thread divergence, reducing data movements and memory footprint of the algorithm, which have been successful strategies to optimize algorithms for GPUs [11], [12]
The proposed algorithm can deal with deviated particle trajectories by a magnetic field. b) We introduce a parallel version for the decoding of the raw input data, which ensures coalesced data write patterns and produces a sorted structure of arrays (SoA) data structure, beneficial to our tracking algorithm. c) We investigate the impact of our algorithm configuration on the physics quality of the results and analyze its computing performance on a variety of GPUs and CPUs
Summary
High-energy physics experiments produce large data streams that must be processed, filtered, and analyzed. Declara et al.: Parallel-Computing Algorithm for High-Energy Physics Particle Tracking and Decoding and accelerators are being tested in different trigger systems [5]–[7]. This is an indication that systems requiring high-throughput can be met in such alternative architectures. The main contributions of this paper are as follows: a) We present a fast tracking algorithm for high-energy physics detectors targeting SIMD architectures called Compass.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
