Narrow Results

Deep neural networks (DNNs) have witnessed widespread adoption across various domains. However, their computational demands pose significant challenges due to the extensive inter-neuron communication within the network. Moreover, the energy consumption of DNNs is substantial, primarily driven by the vast data movement and computational requirements. To overcome these challenges, novel accelerator architectures are essential. In this study, we present a novel heuristic algorithm for neuron grouping, catering to both fully connected and partially pruned DNN models. Our algorithm aims to minimize the overall data communication cost among neuron groups while also considering computational load balance. It outperforms existing heuristic neuron grouping methods classified into three main approaches from the literature by an average improvement in communication cost ranging from 33.01% to 47.11%. By optimizing neuron grouping, our approach may be used to enhance the efficiency of DNN accelerators, enabling improved performance and reduced energy consumption.

In this paper, we present OmpSs, a programming model based on OpenMP and StarSs, that can also incorporate the use of OpenCL or CUDA kernels. We evaluate the proposal on different architectures, SMP, GPUs, and hybrid SMP/GPU environments, showing the wide usefulness of the approach. The evaluation is done with six different benchmarks, Matrix Multiply, BlackScholes, Perlin Noise, Julia Set, PBPI and FixedGrid. We compare the results obtained with the execution of the same benchmarks written in OpenCL or OpenMP, on the same architectures. The results show that OmpSs greatly outperforms both environments. With the use of OmpSs the programming environment is more flexible than traditional approaches to exploit multiple accelerators, and due to the simplicity of the annotations, it increases programmer's productivity.

High-energy and high-beam power accelerators are extensively used for the neutrino physics research. At present, the leading operational facilities are the J-PARC facility in Japan, which recently approached 0.5 MW of the 30 GeV proton beam power, and the Fermilab Main Injector complex that delivers over 0.75 MW of 120 GeV protons on the neutrino target. Besides such type of neutrino superbeams, the concept of neutrino factories offers great promise for particle physics research. In this brief review we present the status and planned upgrades of the J-PARC and Fermilab accelerators and leading proposals for the next generation accelerator-based facilities of both types, their challenges and required and ongoing accelerator R&D programs aimed to address corresponding performance and cost risks.

Thorium is the primary candidate fuel in Thorium-based Molten Salt Reactor (TMSR) and Accelerator Driven System (ADS). However, to utilize Thorium in the reactor and accelerator systems, nuclear data in the literature are not sufficient for neutron-induced fission reaction process. Therefore, we estimated neutron fission cross-section $\sigma$ (mb), independent $Y(A)$ and primary fission fragment yield $Y_I(A)$, mass-dependent average neutron multiplicity $\bar{\nu }(A)$, energy-dependent average neutron multiplicity $\bar{\nu }(E)$ and neutron emission multiplicity distribution $\nu$ of ${}^{232}\mathrm{\text{Th}}$ for different neutron incident energy values, 1 MeV, 1.5 MeV, 3 MeV, 5.5 MeV, 7.7 MeV, 10 MeV, 14.8 MeV and 20 MeV, where the energy region is important for the reactor and accelerator systems used in the industry. The obtained results are compared with experimental data obtained from the literature and are discussed for each energy value.

This paper reviews 40 years of operation of the CERN Proton Synchrotron Booster, from the commissioning in the early 1970s until today's operation as injector for the Large Hadron Collider (LHC) and as one of the backbone accelerators for CERN's fixed-target physics program.

The Large Hadron Collider (LHC) was first suggested (in a documented way) in 1983 (S. Myers and W. Schnell, Preliminary performance estimates for a LEP proton collider, LEP Note 440, April 1983) as a possible future hadron collider to be installed in the 27 km "LEP" tunnel. More than 30 years later the collider has been operated successfully with beam for three years with spectacular performance and has discovered the long-sought-after Higgs boson. The LHC is the world's largest and most energetic particle collider. It took many years to plan and build this large complex machine which promises exciting, new physics results for many years to come.

I describe the LHC design objectives, review some of the more relevant beam effects, define the major accelerator components and parameters, and finally give an overview of the commissioning and operational performance since the initial turn on of the collider. The latter will include the major accident which took place in September 2008 and the subsequent repair and redesign of the faulty components.

The first attempt to circulate beam in the LHC in September 2008 were initially very successful. However, after only nine days of preliminary beam commissioning, on 19 September 2008, disaster struck: the last octant was being ramped up in preparation for high energy operation when a magnet interconnect failed and the enormous energy stored in the superconducting magnets was released in an uncontrolled way and damaged around 600 m of the LHC installed equipment. The next 14 months were crucial for the machine. A crash programme for the repair of the damaged sector was initiated as well as in depth studies to understand the cause of the failure and make design changes which would ensure that such an accident could never reoccur in the future.

In this paper, the story of the four years of the intensive activity since the accident is described starting with the repair of the damaged area and followed by the three very successful years of beam operation.

The TOTEM Experiment is designed to measure the total proton–proton cross-section with the luminosity-independent method and to study elastic and diffractive pp scattering at the LHC. To achieve optimum forward coverage for charged particles emitted by the pp collisions in the interaction point IP5, two tracking telescopes, T1 and T2, are installed on each side of the IP in the pseudorapidity region 3.1≤|η|≤6.5, and special movable beam-pipe insertions — called Roman Pots (RP) — are placed at distances of ±147 m and ±220 m from IP5. This article describes in detail the working of the TOTEM detector to produce physics results in the first three years of operation and data taking at the LHC.

Low energy electrons in accelerators are known to interact with the circulating beam, giving rise to the formation of a so-called e^- cloud. Such e^- cloud may induce detrimental effects on the accelerated beam quality and stability. Those effects have been observed in most accelerators of positively charged particles. A longstanding effort has been so far devoted to understand in detail the physical origin of such e^- cloud, its build-up and its interaction with the circulating beam. We will first describe the origin and the basic features causing e^- cloud formation in accelerators, then we review some of the theoretical work produced to simulate and analyze such phenomenon. In selected cases, theoretical expectations and experimental observations will be compared, to address the importance of benchmarking codes versus observations to reach the required predictive capability. To this scope, codes need realistic input parameters which correctly describe material and surface properties at the basis of such e^- cloud formation and build-up. The experimental efforts, performed worldwide in many surface and material science laboratories, to measure such essential parameters will then be presented and critically reviewed. Finally, we will describe some of the e^- cloud mitigation strategies adopted so far and draw some conclusions.

High energy hadron colliders have been in the forefront of particle physics for more than three decades. At present, international particle physics community considers several options for a 100 TeV proton–proton collider as a possible post-LHC energy frontier facility. The method of colliding beams has not fully exhausted its potential but has slowed down considerably in its progress. This paper briefly reviews the accelerator physics and technology challenges of the future very high energy colliders and outlines the areas of required research and development towards their technical and financial feasibility.

IHEP (the Institute of High Energy Physics, Beijing, China) has started the R&D of high field accelerator magnet technology from 2014 for recently proposed CEPC-SppC (Circular Electron Positron Collider, Super proton–proton Collider) project. The conceptual design study of a 20-T dipole magnet is ongoing with the common coil configuration, and a 12-T model magnet will be fabricated in the next two years. A 3-step R&D process has been proposed to realize this 12-T common-coil model magnet: first, a 12-T subscale magnet will be fabricated with Nb₃Sn and NbTi superconductors to investigate the fabrication process and characteristics of Nb₃Sn coils, then a 12-T subscale magnet will be fabricated with only Nb₃Sn superconductors to test the stress management method and quench protection method of Nb₃Sn coils; the final step is fabricating the 12-T common-coil dipole magnet with HTS (YBCO) and Nb₃Sn superconductors to test the field optimization method of the HTS and Nb₃Sn coils. The characteristics of these R&D steps will be introduced in the paper.

Here we present a short summary of the “Workshop on Beam Acceleration in Crystals and Nanostructures” which has taken place at Fermilab on June 24–25, 2019.

The future of the world-wide HEP community critically depends on the feasibility of the concepts for the post-LHC Higgs factories and energy frontier future colliders. Here we overview the accelerator options based on traditional technologies and consider the need for plasma colliders, particularly, muon crystal circular colliders. We briefly address the ultimate energy reach of such accelerators, their advantages, disadvantages and limits in the view of perspectives for the far future of the accelerator-based particle physics and outline possible directions of R&D to address the most critical issues.

The paper presents a short overview of main physics phenomena which determine how the beam can be accelerated in plasma, crystal or a nanotube.

Crystals were used at Fermilab accelerators for slow extraction and halo collimation in the Tevatron collider, and for channeling radiation generation experiments at the FAST electron linac facility. Here we overview past experience and major outcomes of these studies and discuss opportunities for new crystal acceleration R&D program.

An expert system that acts as an intelligent assistant to operators tuning a particle beam accelerator was developed. The system incorporates three approaches to tuning: (1) Duplicating within a software program the reasoning and the procedures used by an operator to tune an accelerator. This approach has been used to steer particle beams through the transport section of Lawrence Livermore National Laboratory's Advanced Test Accelerator and through the injector section of the Experimental Test Accelerator. (2) Using a model to simulate the position of a beam in an accelerator. The simulation is based on data taken directly from the accelerator while it is running. This approach will ultimately be used by operators of the Experimental Test Accelerator to first compare actual and simulated beam performance in real time, then to determine which set of parameters is optimum in terms of centering the beam, and finally to feed those parameters to the accelerator. Operators can also use the model to determine if a component has failed. (3) Using a mouse to manually select and control the magnets that steer the beam. Operators on the Experimental Test Accelerator can also use the mouse to call up windows that display the horizontal and vertical positions of the beam as well as its current.

Intermediaries such as accelerators support entrepreneurial activity in developing countries by connecting entrepreneurs to critical resources and by reshaping the entrepreneurial ventures so they can better participate in larger markets. Existing research has examined the activities intermediaries undertake and how these activities influence intermediary effectiveness. However, we know much less about which entrepreneurial ventures benefit from intermediation. Using 24 months of pre- and post-intervention sales data for 139 ventures working with a business accelerator in Central America, we find that facilitating resource acquisition is less important than the constraints to change within the entrepreneurial ventures themselves. Thus, our study suggests that although facilitating resource acquisition through venture acceleration is important, it may be insufficient for increasing venture growth. Rather, the malleability of the venture may play a more important role in intermediation effectiveness.

This paper explores the front end innovation activities in a multinational Global Healthcare Company (GHC). A questionnaire was designed and distributed to front end innovators from 20 operating companies to understand team composition, essential skill sets, and the methodology used to assess customer needs, generate ideas, and define the selection criteria used during the go/no go decision. For each category the current state and best practice (based on the views of the individual respondents) for front end innovation was analyzed and reported in this paper. The results of this study reflect the opinions and suggestions of 37 respondents (23 that were at directorship level or above). These results were amalgamated and used to develop existing models of innovation to produce a framework (termed the Total Front End Framework (TFEF)) that should be useful for both senior managers and innovation teams to optimize their activities and increase the efficiency of the innovation process.

In this paper, standing back — looking from afar — and adopting a historical perspective, the field of accelerator science is examined. The subjects explored are: how it grew, what were the forces that made it what it is, where it is now, and what it is likely to be in the future. Clearly, many personal opinions, are offered in this process.

Particle accelerators were initially developed to address specific scientific research goals, yet they were used for practical applications, particularly medical applications, within a few years of their invention. The cyclotron's potential for producing beams for cancer therapy and medical radioisotope production was realized with the early Lawrence cyclotrons and has continued with their more technically advanced successors — synchrocyclotrons, sector-focused cyclotrons and superconducting cyclotrons. While a variety of other accelerator technologies were developed to achieve today's high energy particles, this article will chronicle the development of one type of accelerator — the cyclotron, and its medical applications. These medical and industrial applications eventually led to the commercial manufacture of both small and large cyclotrons and facilities specifically designed for applications other than scientific research.

Medical applications represent the vast majority of the uses for radiotracers. This review addresses how accelerators are employed for the production of high purity radionuclides that are used in basic biomedical research, as well as for clinical medicine both for diagnosing disease and for treatment.