Performance Computing Systems Research Articles

Storage Class Memory Based Hybrid Memory System for Practical Remote Sensing

Koo, S.M.; Seo, J.M.; Song. Y.J., and Baek, S.J., 2020. Storage class memory based hybrid memory system for practical remote sensing. In: Jung, H.-S.; Lee, S., and Ryu, J.-H. (eds.), In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 254-260. Coconut Creek (Florida), ISSN 0749-0208.In the remote sensing domain, large-sized, files such as high-resolution satellite images or sonar video clips from unmanned underwater vehicles are very common. For processing big files, a large main memory is necessary. The main memory capacity highly influences the performance of computer systems. The demand for DRAM capacity has never been satisfied, and more importantly, large DRAM systems suffer from significant power consumption. To ease the problem, a promising solution is to build a hybrid main memory (HMM) system composed of a small number of fast DRAMs and many inexpensive devices. By mimicking large and fast main memory capacity, HMM allows computer systems to run applications that require more DRAM than is installed on the system. In this paper, a novel HMM management scheme for a storage class memory (SCM)-based HMM was introduced. As all the data stored on SCM are already non-volatile, the overall performance of the computer system is enhanced further by not flushing them periodically. The proposed idea was implemented on Linux and its performance was measured using an SCM emulation system. It was shown that HMM efficiently improves performance by up to 77.9 %, compared with a conventional operating system. Additionally, it was demonstrated that the proposed idea's fault recovery mechanisms could restore dirty data that are not yet synchronized with storage, within 91 % of the test time.

Efficiency of UAV-based last-mile delivery under congestion in low-altitude air

Emerging unmanned aerial vehicle (UAV) technologies have motivated logistics carriers to seek last-mile parcel delivery through the air so as to benefit from its convenience and flexibility. However, UAV-based delivery services are limited by several binding factors, such as low battery capacities and short delivery range, which in turn require simultaneous use of a large fleet for commercial scale operations. In such cases, congestion in low-altitude air will inevitably arise. This paper investigates self-organized UAV traffic flow in low altitude 3D airspace, and formulates the user equilibrium condition as a set of partial differential equations. We propose a finite element scheme to numerically solve the traffic equilibrium and compute system performance. Two specific test scenarios for last-mile freight delivery systems are studied, including one with a conventional ground-based distribution facility, and the other with a novel concept of airborne fulfillment center. We evaluate the operational cost and energy consumption of these systems under a variety of system configurations. The results provide insights that could be useful for logistic carriers and policy makers to achieve efficiency and sustainability for last-mile delivery.

Performance Evaluation of Web Servers using Response Time and Bandwidth

Computer system performance can be measured using the amount of useful work accomplished by that system when compared to the time and resources utilized. Useful work here means how well the computer is doing the work it is supposed to do. Most websites work under the support of web servers that usually include hardware (for instance CPU, RAM, Disk, and network) as well as software (web services). The alarming growth in web traffic has led to performance problems and has necessitated much research activities as ways to improve web server performance. In this paper, a web server performance evaluation system that evaluates a server based on time of file execution and bandwidth has been designed and developed. Access log data sets from University of Port Harcourt web site were used to study the system and develop software for checking time of server activity in any domain where the application is executed. The evaluation system so designed and developed can be deployed in any server and the values generated can be used in making useful decisions that will mitigate certain occurrences in the future and also offers site owners information that will be useful in forecasting users visiting behaviours and the effects on the server downtime.

STFL-DDR: Improving the Energy-Efficiency of Memory Interface

Power dissipation is a significant problem limiting the performance of today's computer systems. One of the main contributors to power consumption in microprocessors is data movement in cache and memory interface. Several solutions such as low power interconnects, energy-aware data encoding, and low power signaling have been proposed to mitigate this problem. Almost all of these techniques result in a significant system performance degradation. This article examines the application of a novel technique, called STFL-DDR, for hybrid signaling on low-power DRAM interface. To keep the power consumption low, STFL-DDR employs a high-performance clock rate for transferring data on low power wires. To avoid any signal deterioration, STFL-DDR employs data encoding/decoding to prevent each wire from switching in any two consecutive cycles. STFL-DDR creates new opportunities for optimizing the energy-efficiency of DRAM systems. We compare the efficiency of STFL-DDR with the state-of-the-art methods by simulating a mix of 12 parallel benchmark applications on a muticore system. Our simulation results indicate that STFL can reduce the energy consumption of a contemporary DRAM interface by 17 percent as compared to an LPDDR baseline while achieving the throughput of a high-performance DRAM. Applying STFL to both last level cache and DRAM interface results in improving the system energy, energy-delay product, and performance by 8, 15, and 9 percent respectively. Compared with a high-performance memory interface, STFL improves the system energy and energy-delay product by 25 and 75 percent, while reaching 98 percent of the average performance of the high-performance system.

Distributed MPC for Large Freeway Networks Using Alternating Optimization

The Model Predictive Control (MPC) framework has shown great potential for the control of Variable Speed Limits (VSLs) and Ramp Metering (RM) installations. However, the implementation to large freeway networks remains challenging. One major reason is that, by considering the VSLs to be discrete decision variables, an extremely difficult Mixed Integer Nonlinear Programming (MINLP) optimization problem has to be solved within every controller sampling interval. Consequently, many related papers relax the MINLP problems by considering the VSLs to be continuous variables. This paper proposes two novel MPC algorithms for coordinated control of discrete VSLs and continuous RM rates that do not make this relaxation. The proposed algorithms use a distributed control architecture and an alternating optimization scheme to relax the MINLP optimization problems but still consider the VSLs as discrete variables and, hence, offer a trade-off between computational complexity and system performance. The performance of the proposed algorithms is evaluated in a case study. The case study shows that relaxing the VSLs to be continuous variables with a distributed architecture results in a significant performance loss. Furthermore, both proposed algorithms have a lower computational complexity than the more conventional centralized approach and, as a result, they do manage to solve all optimization problems within the sampling intervals. Moreover, one of the proposed algorithms has a system performance that is remarkably similar to the optimal performance of the centralized approach.

Application of Cluster Analysis with Fuzzy Logic Elements for Ground Environment Assessment of Robotic Group

Emergency situations, that cause risks for human life and health, dictate elevated requirements to completeness and accuracy of the presentation of information about current ground environment. Modern robotic systems include sensors, that operate on different physical principles. This causes incrementation of information entering control system. Computing resources and technical capabilities of robotic systems are limited in range and detection probabilities of appearing objects. In case of insufficient performance of the on-board computer system and high uncertainties of ground environment, robotic systems are not able to perform without combining information from robotic group and producing a single view of ground environment. Complex information from a group of robotic systems occurs in real time and a non-deterministic environment.
 To solve the problem of identifying attribute vectors related to a single object, as well as to evaluate the effectiveness of obtained solutions, is possible using known formulas of the theory of statistical hypothesis testing and probability theory only under the normal distribution law with the known mathematical expectation of an attribute vector and a correlation matrix. However, these conditions are usually not met in practice. Problems also arise when methods of nonparametric statistics are used with an unknown law of probability distribution.
 The new method of identifying attribute vectors is proposed, that does not rely on a statistical approach and, therefore, does not require knowledge of the type of distribution law and the values of its parameters. Proposed method is based on the idea of combining cluster analysis and fuzzy logic, and is relatively simple to the basic methods of multidimensional nonparametric statistics.
 The results of modeling information processes are presented. The advantages of proposed method are shown. The comparative values for the number of false recognitions are given. The recommendations are given for constructing fuzzy inference rules when creating an expert system knowledge base.

RETRACTED: Intelligent edge computing based on machine learning for smart city

To alleviate the huge computing pressure caused by the single mobile edge server computing mode as the amount of data increases, in this research, we propose a method to conduct calculations in a collaborative way. First, the method needs to consider how to encourage devices to cooperate when they are selfish. Second, the method answers the following question: how can collaborative computing be carried out when the device has the intention to cooperate? For example, how can calculations be conducted when there are extensibility and privacy problems in machine learning tasks? In view of the above challenges, a mobile edge server is taken as the focus, and the available resources around the mobile edge server are used for collaborative computing to further improve the computing performance of a mobile edge computing (MEC) system. The alternating direction multiplier method is used to solve the problem. First, the relevant techniques and theories of MEC, Stackelberg principle-subordinate game theory, and the alternating direction method of multipliers (ADMM) are introduced. Then, the problem description and model construction of distributed task scheduling in MEC and machine learning task-based device coordination computing are introduced, and machine learning is applied in the distributed task scheduling algorithm and distributed device coordination algorithm. Finally, the distributed task scheduling algorithm and distributed device coordination algorithm are tested by experiments.

Impact of coupling topology upon noise robustness of small optical reservoirs

In this work, we perform the numerical investigation of the performance of the small optical reservoir computing (RC) systems with four neurons using the commercial software for optical fiber communication system. The small optical RC system consists of the components of the optical fiber communication. The nonlinear function which is required in RC is provided by the erbium-doped optical fiber amplifiers (EDFA). We demonstrate that the EDFA should be operated in the saturated or non-linear regime to obtain a better performance of the small optical RC system. The performance of the small optical RC systems for different topological neuron structures is investigated. The results show that the interconnection between the neurons could offer a better performance than the systems without interconnection between the neurons. Moreover, the input signals with different noise levels are launched into the systems. The results show that the small optical RC system can classify the noisy input optical waveforms even when the signal-to-noise ratio is as low as − 2.55 dB.

Proactive Rebalancing and Speed-Up Techniques for On-Demand High Capacity Ridesourcing Services

We present a probabilistic proactive rebalancing method and speed-up techniques for improving the performance of a state-of-the-art real-time high-capacity fleet management framework. We improve on both computational efficiency and system performance. The speed-up techniques include search-space pruning and I/O cost reduction for parallelization, reducing the computation time by up to 97.67%, in experiments on taxi trips in New York City. The proactive rebalancing routes idle vehicles to future demands based on probabilistic estimates from historical demand, increasing the service rate by 4.8% on average, and decreasing the waiting time and total delay by 5.0% and 10.7% on average, respectively.

Performance and energy consumption of HPC workloads on a cluster based on Arm ThunderX2 CPU

In this paper, we analyze the performance and energy consumption of an Arm-based high-performance computing (HPC) system developed within the European project Mont-Blanc 3. This system, called Dibona, has been integrated by ATOS/Bull, and it is powered by the latest Marvell’s CPU, ThunderX2. This CPU is the same one that powers the Astra supercomputer, the first Arm-based supercomputer entering the Top500 in November 2018. We study from micro-benchmarks up to large production codes. We include an interdisciplinary evaluation of three scientific applications (a finite-element fluid dynamics code, a smoothed particle hydrodynamics code, and a lattice Boltzmann code) and the Graph 500 benchmark, focusing on parallel and energy efficiency as well as studying their scalability up to thousands of Armv8 cores. For comparison, we run the same tests on state-of-the-art x86 nodes included in Dibona and the Tier-0 supercomputer MareNostrum4. Our experiments show that the ThunderX2 has a 25% lower performance on average, mainly due to its small vector unit yet somewhat compensated by its 30% wider links between the CPU and the main memory. We found that the software ecosystem of the Armv8 architecture is comparable to the one available for Intel. Our results also show that ThunderX2 delivers similar or better energy-to-solution and scalability, proving that Arm-based chips are legitimate contenders in the market of next-generation HPC systems.

Food Intake Actions Detection: An Improved Algorithm Toward Real-Time Analysis.

With the increase in life expectancy, one of the most important topic for scientific research, especially for the elderly, is good nutrition. In particular, with an advanced age and health issues because disorders such as Alzheimer and dementia, monitoring the subjects’ dietary habits to avoid excessive or poor nutrition is a critical role. Starting from an application aiming to monitor the food intake actions of people during a meal, already shown in a previously published paper, the present work describes some improvements that are able to make the application work in real time. The considered solution exploits the Kinect v1 device that can be installed on the ceiling, in a top-down view in an effort to preserve privacy of the subjects. The food intake actions are estimated from the analysis of depth frames. The innovations introduced in this document are related to the automatic identification of the initial and final frame for the detection of food intake actions, and to the strong revision of the procedure to identify food intake actions with respect to the original work, in order to optimize the performance of the algorithm. Evaluation of the computational effort and system performance compared to the previous version of the application has demonstrated a possible real-time applicability of the solution presented in this document.

Crowdsourcing human-based computation for medical image analysis: A systematic literature review

Computer-assisted algorithms for the analysis of medical images require human interactions to achieve satisfying results. Human-based computation and crowdsourcing offer a solution to this problem. We performed a systematic literature review of studies on crowdsourcing human-based computation for medical image analysis based on the guidelines proposed by Kitchenham and Charters. We identified 43 studies relevant to the objective of this research. We determined three primary purposes and problems that crowdsourcing human-based computation systems can solve. We found that the users provided five information types. We compared systems that use pre-, post-evaluation and quality control methods to select and filter the user inputs. We analyzed the metrics used for the evaluation of the crowdsourcing human-based computation system performance. Finally, we identified the most popular crowdsourcing human-based computation platforms with their advantages and disadvantages.Crowdsourcing human-based computation systems can successfully solve medical image analysis problems. However, the application of crowdsourcing human-based computation systems in this research area is still limited and more studies should be conducted to obtain generalizable results. We provided guidelines to practitioners and researchers based on the results obtained in this research.

Performance analysis of a flexible flow shop with random and state-dependent batch transport

In many manufacturing contexts, performance modelling of an integrated production and material handling system is a complex problem. Existing research lacks an in-depth consideration of the integration of these two areas. A flexible flow shop with random and state-dependent batch transport, where the batch size depends on the number of jobs in the buffers and the capacity of automated guided vehicles, is considered and modelled as an open queueing network with blocking. A decomposition method of state space is proposed for computing system performance measures. The accuracy and efficiency of the proposed method are demonstrated by comparing the results with simulations from numerical experiments. Meanwhile, the properties of the system, especially for material handling processes, are investigated and analyzed according to the experiments. The results of this paper can be used as a basis for system design, analysis, and resource planning.

High Performance Computing in Resource Poor Settings: An Approach based on Volunteer Computing

High Performance Computing (HPC) systems aim to solve complex computing problems (in a short amount of time) that are either too large for standard computers or would take too long. They are used to solve computational problems in many fields such as medical science (for drug discovery, breast cancer detection in images, etc.), climate science, physics, mathematical science, etc. Existing solutions such as HPC Supercomputer, HPC Cluster, HPC Cloud or HPC Grid are not adapted for resource poor settings (mainly for developing countries) because their fees are generally beyond the funding (particularly for academics) and the administrative complexity to access to HPC Grid creates a higher barrier. This paper presents an approach allowing to build a Volunteer Computing system for HPC in resource poor settings. This solution does not require any additional investment in hardware, but relies instead on voluntary machines already owned by the private users. The experiment has been made on the mathematical problem of solving the matrices multiplication using Volunteer Computing system. Given the success of this experiment, the enrollment of other volunteers has already started. The goal being to create a powerful Volunteer Computing system with the maximum number of computers.

SpackDev: Multi-Package Development with Spack

High Energy Physics (HEP) software environments often involve ∼ hundreds of external packages and libraries, and similar numbers of experiment-specific, science-critical packages—many under continuous development. Managing coherent releases of the external software stack is challenging enough, but managing the highly-collaborative—and distributed—development of a large body of code against such a stack adds even more complexity and room for error. Spack is a popular Python-based package management tool with a specific focus on the needs of High Performance Computing (HPC) systems and system administrators whose strength is orchestrating the discrete download, build, testing, and installation of pre-packaged or tagged third-party software against similarly stable dependencies. As such it is becoming increasingly popular within HEP as that community makes increasing use of HPC facilities, and as efforts to develop future HPC systems utilize Spack to provide scientific software on those platforms [1]. SpackDev is a system to facilitate the simultaneous development of interconnected sets of packages. Intended to handle packages without restriction to one internal build system, SpackDev is integrated with Spack as a command extension in order to leverage features such as dependency calculations and build system configuration, and is generally applicable outside HEP. We describe SpackDev’s features and development over the last two years, initial experience using SpackDev in the context of the LArSoft liquid argon detector toolkit, and work remaining before it can be considered a fully-functional multi-package build system for HEP experiments utilizing Spack.

A Prototype of FPGA Based on Genetic Algorithm Core Connected to a Cluster

In this paper, it was proposed to design and implement a system of parallel processing to find the required solutions from selective algorithms as quickly. Motives for writing paper are the current performance of computer systems that depend on evolutionary algorithms (EA), and the wide spread of the (EA), and its application to a very wide range of scientific fields, and by taking advantage of the Field programmable gate array (FPGA) board due to its high speed of implementation. The values were validated and the Genetic Algorithm (GA) was used as a functional model and implementation. Also, in the most important stages, the process of calculating fitness function, which is considered an executive criterion for the (GA), with terminal computers with high speeds and medium specifications, was done for the purposes of calculating fitness function independently of Board. Identical results were obtained at 100% accuracy by applying the work to a non-linear quadratic.

Global Attraction of ODE-based Mean Field Models with Hyperexponential Job Sizes

Mean field modeling is a popular approach to assess the performance of large scale computer systems. The evolution of many mean field models is characterized by a set of ordinary differential equations that have a unique fixed point. In order to prove that this unique fixed point corresponds to the limit of the stationary measures of the finite systems, the unique fixed point must be a global attractor. While global attraction was established for various systems in case of exponential job sizes, it is often unclear whether these proof techniques can be generalized to non-exponential job sizes. In this paper we show how simple monotonicity arguments can be used to prove global attraction for a broad class of ordinary differential equations that capture the evolution of mean field models with hyperexponential job sizes. This class includes both existing as well as previously unstudied load balancing schemes and can be used for systems with either finite or infinite buffers. The main novelty of the approach exists in using a Coxian representation for the hyperexponential job sizes and a partial order that is stronger than the componentwise partial order used in the exponential case.

Task-Independent Computational Abilities of Semiconductor Lasers with Delayed Optical Feedback for Reservoir Computing

Reservoir computing has rekindled neuromorphic computing in photonics. One of the simplest technological implementations of reservoir computing consists of a semiconductor laser with delayed optical feedback. In this delay-based scheme, virtual nodes are distributed in time with a certain node distance and form a time-multiplexed network. The information processing performance of a semiconductor laser-based reservoir computing (RC) system is usually analysed by way of testing the laser-based reservoir computer on specific benchmark tasks. In this work, we will illustrate the optimal performance of the system on a chaotic time-series prediction benchmark. However, the goal is to analyse the reservoir’s performance in a task-independent way. This is done by calculating the computational capacity, a measure for the total number of independent calculations that the system can handle. We focus on the dependence of the computational capacity on the specifics of the masking procedure. We find that the computational capacity depends strongly on the virtual node distance with an optimal node spacing of 30 ps. In addition, we show that the computational capacity can be further increased by allowing for a well chosen mismatch between delay and input data sample time.

Non-stationary Antonov self-gravitating layer: analytics and numerics

ABSTRACTLarge-scale instability of gravitating systems plays a key role in collisionless relaxation and in reaching a quasi-stationary state at the early stage of evolution. Advanced high-resolution methods and permanently increasing performance of computational systems allow this phenomenon to be studied by means of computer simulations at a new level. In this paper, an approach to verification and validation of computer codes implementing high-resolution methods is proposed. The approach is based on comparisons of the simulation results with exact non-stationary solutions of the Vlasov–Poisson equations. The evolution of the gravitating layer model is considered as an example of implementation of this approach. A one-parameter family of exact models of a non-stationary gravitating layer is described, and their stability to large-scale disturbances in the linear approximation is analytically studied. Non-linear instability development is computed with the use of the fifth-order conservative semi-Lagrangian WENO scheme.

Cognitive Radio Network Assisted by OFDM With Index Modulation

Orthogonal frequency division multiplexing with index modulation (OFDM-IM) transmits additional information bits through the indices of active subcarriers. In this paper, we explore the potential of an improvement of OFDM-IM by reusing inactive subcarriers to transmit signals for a secondary system. We consider a cognitive radio network assisted by OFDM-IM, where the primary user (PU) transmits its signal to the primary receiver, and the secondary user (SU) senses the idle spectrum and transmits its own signal to the secondary receiver via those idle subcarriers. By either combining the signal transmitted from SU or not, we propose two different detection strategies with different trade-offs between the computational complexity and system performance. An asymptotically tight upper bound on the bit error rate is derived to evaluate the error performance.