High Performance Computing Research Research Articles

Multi-resource scheduling of moldable workflows

Resource scheduling plays a vital role in High-Performance Computing (HPC) systems. Most scheduling research in HPC has focused on only a single type of resource (e.g., computing cores or I/O resource). With the advancement in hardware architectures and the increase in data-intensive HPC applications, there is a need to simultaneously consider a diverse set of resources (e.g., computing cores, cache, memory, I/O, and network resources) in the design of runtime schedulers for improving the overall application performance. In this paper, we study multi-resource scheduling to minimize the makespan of computational workflows comprised of moldable parallel jobs. Moldable jobs allow the scheduler to flexibly select a variable set of resources before execution, thus can adapt to the available system resources (as compared to rigid jobs) while staying easy to design and implement (as compared to malleable jobs). We propose a Multi-Resource Scheduling Algorithm (MRSA), which combines a novel resource allocation strategy and an extended list scheduling scheme to schedule the jobs. We prove that, on a system with d types of schedulable resources, MRSA achieves an approximation ratio of 1.619d+2.545d+1 for any d≥1, and a ratio of d+3d23+O(d3) when d is large (i.e., d≥22). We also present improved approximation results for workflows comprised of jobs with special precedence constraints (e.g., series-parallel graphs, trees, and independent jobs). Further, we prove a lower bound of d on the approximation ratio of any list-based scheduling algorithm with local priority considerations. Finally, we conduct a comprehensive set of simulations to evaluate the performance of the algorithm using synthetic workflows of different structures and moldable jobs following different speedup models. The results show that MRSA fares better than the theoretical bound predicts, and that it consistently outperforms two baseline heuristics under a variety of parameter settings, illustrating its robust practical performance.

Fighting COVID-19 with HPC

A special prize for outstanding research in high-performance computing (HPC) applied to COVID-19-related challenges sheds light on the computational science community's efforts to fight the ongoing global crisis.

Inspiring the Next Generation of HPC Engineers with Reconfigurable, Multi-Tenant Resources for Teaching and Research

There is a tradition at our university for teaching and research in High Performance Computing (HPC) systems engineering. With exascale computing on the horizon and a shortage of HPC talent, there is a need for new specialists to secure the future of research computing. Whilst many institutions provide research computing training for users within their particular domain, few offer HPC engineering and infrastructure-related courses, making it difficult for students to acquire these skills. This paper outlines how and why we are training students in HPC systems engineering, including the technologies used in delivering this goal. We demonstrate the potential for a multi-tenant HPC system for education and research, using novel container and cloud-based architecture. This work is supported by our previously published work that uses the latest open-source technologies to create sustainable, fast and flexible turn-key HPC environments with secure access via an HPC portal. The proposed multi-tenant HPC resources can be deployed on a “bare metal” infrastructure or in the cloud. An evaluation of our activities over the last five years is given in terms of recruitment metrics, skills audit feedback from students, and research outputs enabled by the multi-tenant usage of the resource.

Knowledge Structure of the Application of High-Performance Computing: A Co-Word Analysis

As high-performance computing (HPC) plays a key role in the Fourth Industrial Revolution, the application of HPC in various industries is becoming increasingly important. Several studies have reviewed the research trends of HPC but considered only the functional aspects, causing limitations when discussing the application. Thus, this study aims to identify the knowledge structure of the application of HPC, enabling practical and policy support in various industrial fields. Co-word analysis is mainly used to establish the knowledge structure. We first collected 28,941 published papers related to HPC applications and built a co-word network that used author keywords. We performed centrality analysis and cluster analysis of the co-word network; as a result, we derived the major keywords and 18 areas of HPC applications. To validate the knowledge structure, we conducted a case study to find opportunities for HPC research plans in the research community. As a result, we discovered 17 new research topics and presented their research priorities by conducting expert interviews and Analytic Hierarchy Process. The findings of this study contribute to an understanding of the application of HPC, to exploring promising research fields for technological and social development, and to supporting research plans for successful technology commercialization.

Region-specific guidelines to encourage SMEs to use high performance computing

PurposeThis paper aims to explore high performance computing (HPC) in the context of the South East region of Ireland, which hosts a publicly available HPC infrastructure, by identifying whether companies, especially small and medium enterprises (SMEs), are using, or are prepared to use, HPC to improve their business processes, expansion and sustainability. The result of the analysis provides region-specific guidelines that are meant to improve the HPC landscape in the region. The lessons learned from this research may apply to other similar, and developing, European regions.Design/methodology/approachThis paper explores the use of HPC in the context of the South East region of Ireland and examines whether companies, especially SMEs, are benefiting from the use of publicly available HPC infrastructure in the region. This paper also provides a set of recommendations, of a policy nature, and required actions to increase HPC usage, based on the reality of the region. Therefore, the first step in the process was to understand the HPC landscape in the South East region of Ireland. Interviews were conducted with higher education institute (HEI) staff who were knowledgeable about the HPC infrastructure of their institutes and also about whether collaboration between the HEIs and businesses from the same region exists. The interview findings allowed the proposal of region-specific guidelines to improve the HPC landscape and collaboration in the region. The guidelines were analysed and refined in a focus group with key regional stakeholders from academia, industry and government, who have experience and expertise in high-technology transfer processes happening in the region.FindingsThe findings of the current study strongly suggest that HPC usage by SMEs in the South East region of Ireland is still incipient; and that HPC knowledge is currently inadequately transferred from the HEI hosting the HPC infrastructure to public and private sector organisations based in the region. The findings also demonstrate that there are no courses or training programmes available dedicated to HPC and that the level of collaboration between the HEI hosting the HPC infrastructure and industry in the region is minimal as regards HPC usage and projects. Therefore, there is a need to put specific targeted policies and actions, both from a regional government and HEI perspectives, in place to encourage SMEs to optimise their processes by using HPC.Originality/valueThis research is unique as it provides customised region-specific recommendations (RSR) and feasible actions to encourage industry, especially SMEs, to use HPC and collaborate around it. The literature review identified that there is a lack of studies that can inform policymakers to include HPC in their innovation agenda. Previous research studies specifically focussing on HPC policies are even more scarce. Most of the existing research pertaining to HPC focusses on the technical aspect of HPC; therefore, this research and paper bring a new dimension to existing HPC research. Even though this research was focussed on the South East of Ireland region, the model that generated the RSRs can be extrapolated and applied to other regions that need to develop their HPC landscape and the use of HPC among SMEs in their respective regions.

The MVAPICH project: Transforming research into high-performance MPI library for HPC community

High-Performance Computing (HPC) research, from hardware and software to the end applications, provides remarkable computing power to help scientists solve complex problems in science, engineering, or even daily business. Over the last decades, Message Passing Interface (MPI) libraries have been powering numerous scientific applications, such as weather forecasting, earthquake simulations, physic, and chemical simulations, to conduct ever-large-scale experiments. However, it is always challenging to practice HPC research. In this article, we use the well-established MVAPICH project to discuss how researchers can transform the HPC research into a high-performance MPI library and translate it to the HPC community.

PREFACE

It is with great pleasure that I introduce the proceedings of the 2019 High Performance Computing and Computational Intelligence Conference (HPCCI 2019) which was held recently Chengdu, China during October 25-27, 2019. HPCCI 2019 was dedicated to technical issues related to High Performance Computing and Computational Intelligence.High performance computing represents the national strength of a country and plays an essential supporting role in the development of national science and technology. Quantum computing, as a new computing method, could be an emerging area of high-performance computing in the future. The development of high performance computing may have some breakthroughs in these areas. At the same time, it is found that the implicit parallelism provided by the quantum system improves the performance of the algorithm. Quantum theory may be an important research direction to construct the basic theoretical system for some intelligence schemes in the future. The development of computational intelligence cannot be separated from the support of high-performance computing. The development of the two subject areas has gradually formed an integrated system. Based on it, the objectives of the conference were not only to provide opportunities to the international scientists to present state of the art research in High Performance Computing and Computational Intelligence but also to provide a forum to discuss ideas for future directions in the field and explore possibilities for collaborative research projects.The conference program included keynote, oral, and poster presentations from scholars working in the areas of High Performance Computing and Computational Intelligence from all over the world. It covered recent trends and progress made in the field of High Performance Computing and Computational Intelligence. Keynote lectures regarding the latest information in their respective areas of expertise were shared by experts.The proceedings present a selection of high quality papers submitted to the conference by researchers from universities, research institutes, and industry. All papers were subjected to peer-review by conference committee members and international reviewers. The papers were selected based on their quality and their relevance to the conference. The proceedings present recent advances in the fields of High performance Computing and Computing Intelligence.I would like to express special gratitude to members of the conference committee and organizers of the conference. I would also like to thank the reviewers for their valuable time and advice which helped in improving the quality of the papers selected for presentation at the conference and for publication in the proceedings. Finally I want to thank the authors, the members of the organizing committee, the reviewers, the chairpersons, sponsors, and all other conference participants for their support of HPCCI 2019.Conference ChairProf. Peng Wang Southwest Minzu University, China

Reducing energy usage in resource-intensive Java-based scientific applications via micro-benchmark based code refactorings

In-silico research has grown considerably. Today?s scientific code involves long-running computer simulations and hence powerful computing infrastructures are needed. Traditionally, research in high-performance computing has focused on executing code as fast as possible, while energy has been recently recognized as another goal to consider. Yet, energy-driven research has mostly focused on the hardware and middleware layers, but few efforts target the application level, where many energy-aware optimizations are possible. We revisit a catalog of Java primitives commonly used in OO scientific programming, or micro-benchmarks, to identify energy-friendly versions of the same primitive. We then apply the micro-benchmarks to classical scientific application kernels and machine learning algorithms for both single-thread and multi-thread implementations on a server. Energy usage reductions at the micro-benchmark level are substantial, while for applications obtained reductions range from 3.90% to 99.18%.

Gathering requirements for advancing simulations in HPC infrastructures via science gateways

Compute-intensive simulations are often based on complex scientific theories and necessitate high-performance computing (HPC) infrastructures to deliver results in reasonable time. While domain researchers are experts in their field and apply sophisticated theoretical models in computational simulations, they are not necessarily also HPC experts or IT specialists in general. Thus, they appreciate easy-to-use solutions tailored to their research, which hide the complex underlying computing and data infrastructures. Science gateways form such end-to-end solutions and their development for compute-intensive simulations necessitates expertise to connect HPC research infrastructures including grid and cloud infrastructures to support with the efficient access to such resources. HPC experts and IT specialists fulfilling this task may have only rudimentary knowledge about the research domain of a simulation. Thus, it is crucial that they gather the requirements of a research use case, which they aim to support efficiently via a science gateway.In the last 10 years quite a few web development frameworks, science gateway frameworks and APIs with different foci and strengths have evolved to support the developers of science gateways in implementing an intuitive solution for a target research domain. The selection of a suitable technology for a specific use case is essential and helps reducing the effort in implementing the science gateway by re-using existing software or frameworks. Thus, a solution for a user community can be provided more efficiently. This paper introduces the general architecture of science gateways, goes into detail for criteria to design science gateways efficiently and gives examples of mature science gateways and science gateway frameworks.

DESIGN OF RESEARCH-SPECIFIC CLOUD SYSTEM FOR HIGH-PERFORMANCE COMPUTING OVER CLOUD SERVICE

Traditional high-performance computing (HPC) services for scientific applications offer high performance and advantages such as large-scale task execution, concurrency, load balancing, and system utilization. However, they fail to provide any scalability due to their inflexible architectures. Therefore, to provide scalability as well as cloud advantages, “HPC over Cloud services” has become a major area of interest in the field of HPC research. To provide a HPC over Cloud service, any of three representative base architectures can be adopted. These are bare metal (BM)-based, container-based, and virtual machine (VM)-based. The results of a study which compared these three base architectures indicated that the container- and BM-based architectures offer almost the same levels of performance. In this study, we set out to design a research-specific cloud system for HPC over Cloud service. As part of our efforts, we build a test-bed system which provides a real container-based HPC over Cloud service for evaluation. The results of our experiments indicate that the proposed system architecture for a container-based HPC over Cloud service can be successfully adopted for practical applications.

High‐performance computing: to boldly go where no human has gone before

Computing and computational science are well on their way to enter into the exascale era while High Performance Computing (HPC) is now present in many spheres and domains of modern society. New technologies such as cloud computing, big data, and connected objects open many perspectives and possibilities that were unattainable until now. These are high times for computing and for HPC in particular. The High Performance Computing \& Simulation conference aims to bring together researchers working on various aspects of HPC, their design and use, and their applications. As has been the conference tradition, selected extended papers of HPCS 2012 are presented in this special issue. These should be interesting contributions that supplement the current state-of-the-art research in HPC. To relate these works and highlight their value, in this article, we briefly trace back the history of HPC, sketch the state of the present research and development in the field, and project some of the challenges and future trends anticipated.

Developing High Performance Computing Resources for Teaching Cluster and Grid Computing Courses

Abstract High-Performance Computing (HPC) and the ability to process large amounts of data are of paramount importance for UK business and economy as outlined by Rt Hon David Willetts MP at the HPC and Big Data conference in February 2014. However there is a shortage of skills and available training in HPC to prepare and expand the workforce for the HPC and Big Data research and development. Currently, HPC skills are acquired mainly by students and staff taking part in HPC-related research projects, MSc courses, and at the dedicated training centres such as Edinburgh Universitys EPCC. There are few UK universities teaching the HPC, Clusters and Grid Computing courses at the undergraduate level. To address the issue of skills shortages in the HPC it is essential to provide teaching and training as part of both postgraduate and undergraduate courses. The design and development of such courses is challenging since the technologies and software in the fields of large scale distributed systems such as Cluster, Cloud and Grid computing are undergoing continuous change. The students completing the HPC courses should be proficient in these evolving technologies and equipped with practical and theoretical skills for future jobs in this fast developing area. In this paper we present our experience in developing the HPC, Cluster and Grid modules including a review of existing HPC courses offered at the UK universities. The topics covered in the modules are described, as well as the coursework project based on practical laboratory work. We conclude with an evaluation based on our experience over the last ten years in developing and delivering the HPC modules on the undergraduate courses, with suggestions for future work.

FERMI: the most powerful computational resource for Italian scientists

By the last quarter of 2012 CINECA, the largest super-computing center in IT and one of the most important worldwide, will provide to all the scientific community one of the world fastest supercomputer, the IBM BlueGene-Q, named FERMI to honor the famous Italian physicist. FERMI with its advanced technology and its 2.1 PFlop peak perform, will allow the Italian scientists for the fastest and most advanced technology available worldwide for scientific calculations. CINECA provides High Performance Computing (HPC) infrastructure for Italian and European researchers and operates in the technological transfer sector through high performance scientific computing, the management and development of networks and web based services, and the development of complex information systems for treating large amounts of data. Moreover CINECA has been appointed by the Italian Ministry of Research and University to represent IT in the European HPC research infrastructure (PRACE). The aim of the HPC Department is to develop and promote technical and scientific services for the Italian and European research community. It constantly invests in Supercomputing technologies, thus being able to provide researchers with some of the hardware resources among the most powerful and energy efficient in Europe. Twice a year CINECA will directly award in excess of 40 million processor core hours, to ensure an adequate supply to scientists and engineers for HPC-related research, and thanks to the new supercomputer the core hours awarded will rise above 1 billion.

Java in the High Performance Computing arena: Research, practice and experience

The rising interest in Java for High Performance Computing (HPC) is based on the appealing features of this language for programming multi-core cluster architectures, particularly the built-in networking and multithreading support, and the continuous increase in Java Virtual Machine (JVM) performance. However, its adoption in this area is being delayed by the lack of analysis of the existing programming options in Java for HPC and thorough and up-to-date evaluations of their performance, as well as the unawareness on current research projects in this field, whose solutions are needed in order to boost the embracement of Java in HPC.This paper analyzes the current state of Java for HPC, both for shared and distributed memory programming, presents related research projects, and finally, evaluates the performance of current Java HPC solutions and research developments on two shared memory environments and two InfiniBand multi-core clusters. The main conclusions are that: (1) the significant interest in Java for HPC has led to the development of numerous projects, although usually quite modest, which may have prevented a higher development of Java in this field; (2) Java can achieve almost similar performance to natively compiled languages, both for sequential and parallel applications, being an alternative for HPC programming; (3) the recent advances in the efficient support of Java communications on shared memory and low-latency networks are bridging the gap between Java and natively compiled applications in HPC. Thus, the good prospects of Java in this area are attracting the attention of both industry and academia, which can take significant advantage of Java adoption in HPC.

Linear Stability of Partitioned Runge–Kutta Methods

We study the linear stability of partitioned Runge–Kutta (PRK) methods applied to linear separable Hamiltonian ODEs and to the semidiscretization of certain Hamiltonian PDEs. We extend the work of Jay and Petzold [Highly Oscillatory Systems and Periodic Stability, Preprint 95-015, Army High Performance Computing Research Center, Stanford, CA, 1995] by presenting simplified expressions of the trace of the stability matrix, ${tr}M_s$, for the Lobatto IIIA–IIIB family of symplectic PRK methods. By making the connection to Padé approximants and continued fractions, we study the asymptotic behavior of ${tr}M_s(\omega)$ as a function of the frequency $\omega$ and stage number s.

Petaflop supercomputers of China

After ten years of development, high performance computing (HPC) in China has made remarkable progress. In November, 2010, the NUDT Tianhe-1A and the Dawning Nebulae respectively claimed the 1st and 3rd places in the Top500 Supercomputers List; this recognizes internationally the level that China has achieved in high performance computer manufacturing. The nation has already deployed and will deploy further supercomputers with computing capacities of hundreds of teraflops to petaflops in Beijing, Shanghai, Shenzhen, Tianjin, Jinan, and Changsha, which will greatly strengthen the supercomputing infrastructure. It should be noted that domestic high performance processors will be used in some of these systems. China has made significant advances in large-scale parallel computing areas such as low-power design, parallel programming, and reliability, as well as large-scale application software. Dedicated to HPC efforts in China, this issue of Frontiers of Computer Science in China consists of papers from most major Chinese HPC vendors and research organizations. There are two survey papers, one of which by Xianghui Xie et al. elaborates on the evolution of supercomputers, while the other by Yunquan Zhang et al. provides a perspective of China’s HPC development by examining China’s HPC Top100 List. Four other papers focus on the technical aspects of three Chinese supercomputers. They cover: the heterogeneous architecture of Tianhe-1, by Xuejun Yang et al.; the design of the controller and communication system of the Dawning 6000, by Ninghui Sun et al.; and the balanced system design in DeepComp, by Mingfa Zhu et al. Finally, Zeyao Mo et al. present their research on JASMIN, a parallel programming infrastructure for scientific computing. Recently, there have been at least three major factors driving the evolution of supercomputers: increasing concern for power and computation efficiency, and reliability, the rapid growth of data-intensive applications, and cloud computing. As a result, much work on HPC focuses on at least two major areas: heterogeneous architectures and general parallel programming models. In China, another driving factor of HPC is the use of domestic hardware and software. With the strong support of the Chinese government and the efforts of the entire HPC community, we confidently believe in a flourishing future for Chinese HPC.

Development of Biological Warfare Sensors Using SGI High Performance Computers

Astract: The Center for Computational Science (CCS) of the US Naval Research Laboratory (NRL) conducts leading edge research in High Performance Computing. CCS currently has two SGI Altix 3700s, one Altix 4700, and one SGI ICE machine. Recently the center has seen an increased interest from scientists at NRL who have been running MatLab™ on their local PC's and workstations but who need more computational power. One such application is the development of biological warfare point detection sensors where time-tosolution for a single run can take over 30 hours to complete and many runs are necessary during the development process. This paper describes the issues that were encountered in the port of this code to the SGI High Performance Computing (HPC) computers at NRL and provides a paradigm for moving other computationally intensive MatLab™ programs to HPC machines.

The quest for petascale computing

Although the challenges to achieving petascale computing within the next decade are daunting, several software and hardware technologies are emerging that could help us reach this goal. The authors review these technologies and consider new algorithms capable of exploiting a petascale computer's architecture. One petaflop per second is a rate of computation corresponding to 10/sup 15/ floating-point operations per second. To be of use in scientific computing, a computer capable of this prodigious speed needs a main memory of tens or hundreds of terabytes and enormous amounts of mass storage. Sophisticated compilers and high memory and I/O bandwidth are also essential to exploit the architecture efficiently. To mask the hardware and software complexities from the scientific end user, it would be advantageous to access and use a petascale computer through an advanced problem-solving environment. Immersive visualization environments could play an important role in analyzing and navigating the output from petascale computations. Thus, petascale computing is capable of driving the next decade of research in high-performance computing and communications and will require advances across all aspects of it.

The Army High Performance Computing Research Center

The US Army High Performance Computing Research Center (AHPCRC) provides nationwide leadership in high performance computing (HPC) research. Its supercomputing resources are maintained and operated by the Minnesota Supercomputer Center, Inc. AHPCRC researchers are developing novel techniques and advanced algorithms and applying them to physical problems, as well as evaluating and using advanced computing platforms for HPC applications. The Center trains future scientists and engineers by sponsoring computer science educational research at the undergraduate, MS, PhD, and postdoctoral levels. It provides a framework for HPC technology transfer and infrastructure support to Army laboratories by conducting joint research projects between Army scientists and Center researchers; by jointly supervising the staff scientists placed by the Center at the Army laboratories; and by encouraging visits between Army scientists and Center researchers. Finally, the AHPCRC provides a state-of-the-art HPC environment to its academic partners, to Army and Defense Department scientists, and to industrial and international collaborators.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Centers of Supercomputing— the University of Minnesota Army High Performance Computing Research Center

Centers of Supercomputing— the University of Minnesota Army High Performance Computing Research Center