70 research outputs found
Large-scale DAQ tests for the LHCb upgrade
No full textThe Data Acquisition (DAQ) of the LHCb experiment[1] will be upgraded in 2020 to a high-bandwidth trigger-less readout system. In the new DAQ event fragments will be forwarded to the to the Event Builder (EB) computing farm at 40 MHz. Therefore the front-end boards will be connected directly to the EB farm through optical links and PCI Express based interface cards. The EB is requested to provide a total network capacity of 32Tb/s, exploiting about 500 nodes. In order to get the required network capacity we are testing various technology and network protocols on large scale clusters. We developed on this purpose an Event Builder implementation designed for an InfiniBand interconnect infrastructure. We present the results of the measurements performed to evaluate throughput and scalability measurements on HPC scale facilities
The Trigger and Data Acquisition System for the KM3NeT-Italia towers
No full textKM3NeT-Italia is an INFN project supported with Italian PON fundings for building the core of the Italian node of the KM3NeT neutrino telescope. The detector, made of 700 10′′ Optical Modules (OMs) lodged along 8 vertical structures called towers, will be deployed starting from fall 2015 at the KM3NeT-Italy site, about 80 km off Capo Passero, Italy, 3500 m deep. The all data to shore approach is used to reduce the complexity of the submarine detector, demanding for an on-line trigger integrated in the data acquisition system running in the shore station, called TriDAS. Due to the large optical background in the sea from 40K decays and bioluminescence, the throughput from the underwater detector can range up to 30 Gbps. This puts strong constraints on the design and performances of the TriDAS and of the related network infrastructure. In this contribution the technology behind the implementation of the TriDAS infrastructure is reviewed, focusing on the relationship between the various components and their performances. The modular design of the TriDAS, which allows for its scalability up to a larger detector than the 8-tower configuration is also discussed
Endoscopic dilation in pediatric esophageal strictures: a literature review
No full textEsophageal strictures in pediatric age are a quite common condition due to different etiologies. Esophageal strictures can be divided in congenital, acquired and functional. Clinical manifestations are similar and when symptoms arise, endoscopic dilation is the treatment of choice. Our aim was to consider the efficacy of this technique in pediatric population, through a wide review of the literature
Deep Neural Networks for energy reconstruction of Inverse Beta Decay events in JUNO
No full textThe Jiangmen Underground Neutrino Observatory (JUNO) is a scintillation detector, currently under construction, which aims to solve the neutrino mass hierarchy by measuring reactor electron antineutrino energy spectrum with a a resolution of 3%/sqrt(E(MeV)) – the highest ever achieved in a large mass neutrino detector. Several approaches for energy reconstruction are being evaluated on simulated data, and Deep Learning methods have already shown promising results, both in accuracy and in efficiency. In this work, a new Convolutional Neural Network with a rotational invariant architecture is trained on a small dataset of 160k instances, and is fine-tuned to exploit the detector’s spherical symmetry and make use of position and timing data from individual photomultipliers. This approach proves to be insensitive to the presence of dark noise from thermal fluctuations, leading to a (2.45+-0.03)% visual energy resolution at 2 MeV, only slightly higher than the 2.2% expected from theory, with a reconstruction bias well below 1%. However, a simpler Fully Connected Neural Network, replicated from previous work, which uses only integral data and is trained on a larger dataset (750k instances), leads to a slightly better resolution of (2.26+-0.05)% at 2 MeV, while being more sensitive to added noise – proving that there could still be some margin of improvement for more complex methods
Deep Neural Networks for energy reconstruction of Inverse Beta Decay events in JUNO
No full textThe Jiangmen Underground Neutrino Observatory (JUNO) is a scintillation detector, currently under construction, which aims to solve the neutrino mass hierarchy by measuring reactor electron antineutrino energy spectrum with a a resolution of 3%/sqrt(E(MeV)) – the highest ever achieved in a large mass neutrino detector. Several approaches for energy reconstruction are being evaluated on simulated data, and Deep Learning methods have already shown promising results, both in accuracy and in efficiency. In this work, a new Convolutional Neural Network with a rotational invariant architecture is trained on a small dataset of 160k instances, and is fine-tuned to exploit the detector’s spherical symmetry and make use of position and timing data from individual photomultipliers. This approach proves to be insensitive to the presence of dark noise from thermal fluctuations, leading to a (2.45+-0.03)% visual energy resolution at 2 MeV, only slightly higher than the 2.2% expected from theory, with a reconstruction bias well below 1%. However, a simpler Fully Connected Neural Network, replicated from previous work, which uses only integral data and is trained on a larger dataset (750k instances), leads to a slightly better resolution of (2.26+-0.05)% at 2 MeV, while being more sensitive to added noise – proving that there could still be some margin of improvement for more complex methods.ope
Endoscopic ultrasound in pediatric population: a comprehensive review of the literature
No full textEndoscopic ultrasonography (EUS) with or without fine needle aspiration/biopsy (FNA/B) is a well-established diagnostic tool in adults for the evaluation and management of gastrointestinal (GI) tract disorders. Its use in children is still limited as well as literature in pediatric age is limited, although the application of EUS is now increasing. The present article aims to review the current literature about EUS indication, accuracy and safety in pediatric age
An integrated infrastructure in support of software development
No full textThis paper describes the design and the current state of implementation of an infrastructure made available to software developers within the Italian National Institute for Nuclear Physics (INFN) to support and facilitate their daily activity. The infrastructure integrates several tools, each providing a well-identified function: project management, version control system, continuous integration, dynamic provisioning of virtual machines, efficiency improvement, knowledge base. When applicable, access to the services is based on the INFN-wide Authentication and Authorization Infrastructure. The system is being installed and progressively made available to INFN users belonging to tens of sites and laboratories and will represent a solid foundation for the software development efforts of the many experiments and projects that see the involvement of the Institute. The infrastructure will be beneficial especially for small- and medium-size collaborations, which often cannot afford the resources, in particular in terms of know-how, needed to set up such services. © Published under licence by IOP Publishing Ltd
Explorations of the Viability of ARM and Xeon Phi for Physics Processing
No full textWe report on our investigations into the viability of the ARM processor and the Intel Xeon Phi co-processor for scientific computing. We describe our experience porting software to these processors and running benchmarks using real physics applications to explore the potential of these processors for production physics processing.We report on our investigations into the viability of the ARM processor and the Intel Xeon Phi co-processor for scientific computing. We describe our experience porting software to these processors and running benchmarks using real physics applications to explore the potential of these processors for production physics processing.We report on our investigations into the viability of the ARM processor and the Intel Xeon Phi co-processor for scientific computing. We describe our experience porting software to these processors and running benchmarks using real physics applications to explore the potential of these processors for production physics processing
Deep Neural Networks for Electron-Positron discrimination in the JUNO experiment
No full textNeutrino physics has always been an important area of research in particle physics, especially since
the discovery of neutrino oscillations. The Jiangmen Underground Neutrino Observatory (JUNO) is a
new large liquid scintillator detector that aims to solve the neutrino mass hierarchy measuring reactor
neutrino interaction in the detector via inverse beta decay. One of the most relevant background
eects is given by the presence of electrons which, even if they don't take part in the decay reaction,
leave a trace very similar to that of positrons in the liquid scintillator. High energies experimental
physics has always had to deal with the management and observation of large amounts of data; for
this reason, nowadays, the development of algorithms and the use of computer techniques are some of
the most important skills that form the background preparation of a physicist. With the advent of the
so-called big data and the rapid development of hardware components, the eld of articial intelligence
has made numerous progresses in recent years. In this thesis the electrons-positrons discrimination in
JUNO experiment are investigated through the use of articial neural networks.
The work is organized by rst introducing the main features of JUNO and the inverse beta decay; then
a brief illustration of the modern deep learning techniques is given. Finally, the data set is presented,
together with the development of the techniques, the data analysis and the obtained results.ope
- …
