We describe a fully GPU-based implementation of the first level trigger for the upgrade of the LHCb detector, due to start data taking in 2021. We demonstrate that our implementation, named Allen, ...can process the 40 Tbit/s data rate of the upgraded LHCb detector and perform a wide variety of pattern recognition tasks. These include finding the trajectories of charged particles, finding proton–proton collision points, identifying particles as hadrons or muons, and finding the displaced decay vertices of long-lived particles. We further demonstrate that Allen can be implemented in around 500 scientific or consumer GPU cards, that it is not I/O bound, and can be operated at the full LHC collision rate of 30 MHz. Allen is the first complete high-throughput GPU trigger proposed for a HEP experiment.
The 2020 upgrade of the LHCb detector will vastly increase the rate of collisions the online system needs to process in software in order to filter events in real-time. 30 million collisions per ...second will pass through a selection chain where each step is executed conditional to its prior acceptance. The Kalman filter is a process of the event reconstruction that, due to its time characteristics and early execution in the selection chain, consumes 40% of the whole reconstruction time in the current trigger software. This makes it a time-critical component as the LHCb trigger evolves into a full software trigger in the upgrade. The algorithm Cross Kalman allows performance tests across a variety of architectures, including multi and many-core platforms, and has been successfully integrated and validated in the LHCb codebase. Since its inception, new hardware architectures have become available exposing features that require fine-grained tuning in order to fully utilize their resources. In this paper we present performance benchmarks and explore the Intel® Skylake and Intel® Knights Landing architectures in depth. We determine the performance gain over previous architectures and show that the efficiency of our implementation is close to the maximum attainable given the mathematical formulation of our problem.
For Run 2 of the LHC, LHCb is replacing a significant part of its event filter farm with new compute nodes. For the evaluation of the best performing solution, we have developed a method to convert ...our high level trigger application into a stand-alone, bootable benchmark image. With additional instrumentation we turned it into a self-optimising benchmark which explores techniques such as late forking, NUMA balancing and optimal number of threads, i.e. it automatically optimises box-level performance. We have run this procedure on a wide range of Haswell-E CPUs and numerous other architectures from both Intel and AMD, including also the latest Intel micro-blade servers. We present results in terms of performance, power consumption, overheads and relative cost.
The LHCb DAQ Network is a real time high performance network, in which 350 data sources send data over a Gigabit Ethernet LAN to more than 1500 receiving nodes. The aggregated throughput of the ...application, called Event Building, is more than 60 Gbps. The protocol employed by LHCb makes the sending nodes transmit simultaneously portions of events to one receiving node at a time, which is selected using a credit-token scheme. The resulting traffic is very bursty and sensitive to irregularities in the temporal distribution of packet-bursts to the same destination or region of the network. In order to study the relevant properties of such a dataflow, a non-disruptive monitoring setup based on a networking capable FPGA (Netfpga) has been deployed. The Netfpga allows order of hundred nano-second precise time-stamping of packets. We study in detail the timing structure of the Event Building communication, and we identify potential effects of micro-bursts like buffer packet drops or jitter.
The Large Hadron Collider beauty (LHCb) experiment at CERN is undergoing an upgrade in preparation for the Run 3 data collection period at the Large Hadron Collider (LHC). As part of this upgrade, ...the trigger is moving to a full software implementation operating at the LHC bunch crossing rate. We present an evaluation of a CPU-based and a GPU-based implementation of the first stage of the high-level trigger. After a detailed comparison, both options are found to be viable. This document summarizes the performance and implementation details of these options, the outcome of which has led to the choice of the GPU-based implementation as the baseline.
The LHCb Data Acquisition during LHC Run 1 Alessio, F; Brarda, L; Bonaccorsi, E ...
Journal of physics. Conference series,
01/2014, Letnik:
513, Številka:
1
Journal Article
Recenzirano
Odprti dostop
The LHCb Data Acquisition system reads data from over 300 read-out boards and distributes them to more than 1500 event-filter servers. It uses a simple push-protocol over Gigabit Ethernet. After ...filtering, the data is consolidated into files for permanent storage using a SAN-based storage system. Since the beginning of data-taking many lessons have been learned and the reliability and robustness of the system has been greatly improved. We report on these changes and improvements, their motivation and how we intend to develop the system for Run 2. We also will report on how we try to optimise the usage of CPU resources during the running of the LHC ("deferred triggering") and the implications on the data acquisition.
The LHCb experiment at CERN is undergoing an upgrade in preparation for the Run 3 data taking period of the LHC. As part of this upgrade the trigger is moving to a fully software implementation ...operating at the LHC bunch crossing rate. We present an evaluation of a CPU-based and a GPU-based implementation of the first stage of the High Level Trigger. After a detailed comparison both options are found to be viable. This document summarizes the performance and implementation details of these options, the outcome of which has led to the choice of the GPU-based implementation as the baseline.
We describe a fully GPU-based implementation of the first level trigger for the upgrade of the LHCb detector, due to start data taking in 2021. We demonstrate that our implementation, named Allen, ...can process the 40 Tbit/s data rate of the upgraded LHCb detector and perform a wide variety of pattern recognition tasks. These include finding the trajectories of charged particles, finding proton-proton collision points, identifying particles as hadrons or muons, and finding the displaced decay vertices of long-lived particles. We further demonstrate that Allen can be implemented in around 500 scientific or consumer GPU cards, that it is not I/O bound, and can be operated at the full LHC collision rate of 30 MHz. Allen is the first complete high-throughput GPU trigger proposed for a HEP experiment.
PDF
