High Performance Cloud Computing Research Articles

Testing Uncertainty in ILUTE - An Integrated Land Use -Transportation Micro-simulation Model of Demographic Updating

In microsimulation models, the behaviours of agents as intelligent objects are modeled through various behavioral modeling techniques, which give rise to uncertainties in the modeling workflow and predicted outputs. ILUTE (Integrated Land Use Transportation Environment model) is an agent-object-based microsimulation model system designed to simulate different activities in a city, in which many agents (persons, families, households, firms, etc.) intelligently and/ or randomly act and interact in a complex way. A 150,000 (150K) synthetic base year (1986) household sample of the Greater Toronto and Hamilton Area (GTHA) population is used as the starting point for undertaking twenty-year historical simulations (1986-2006). In order to analyze the uncertainty associated with stochastic behaviours in the ILUTE model system, a 1000 independent runs using different initial random number seeds are generated and tested through parallel programming and using high performance cloud computing (HPC). In this paper, the behaviour of the demographic updating module is tested by examining the variability and the distribution of its outputs including births, deaths, families, and households across simulation runs. The test results show that all the simulated outputs’ distributions are generally normally, or very close to normally, distributed. In general, ILUTE can simulate demographic updating process with considerable reliability and confidence.

Democratizing digital design and manufacturing using high performance cloud computing: Performance evaluation and benchmarking

Cloud computing is an innovative computing paradigm that can bridge the gap between increasing computing demands in computationally intensive tasks for digital design and manufacturing applications and limited resources, scalability, flexibility, and agility in traditional computing paradigms. In light of the benefits of cloud computing, cloud-based high performance computing (HPC) has the potential to enable users to not only accelerate computationally expensive tasks, but also to reduce costs by utilizing on-demand, ubiquitous, seamless, and user-friendly access to remote engineering application packages as well as remote HPC resources. However, due to uncertainty about computing performance on the cloud, many manufacturers find it challenging to justify and adopt Cloud-Based Design and Manufacturing (CBDM). Therefore, the objective of this research is to evaluate the performance of solving a large-scale engineering problem using finite element analysis on several public HPC clouds as well as introduce a new workflow for CBDM. A set of experiments is conducted to compare the performance of the public HPC clouds with that of a standard workstation and a dedicated in-house supercomputer. The performance metrics include elapsed time, speedup, scalability, and stability. Experimental results have shown that the Azure Cloud with 32 cores and the Nimbix Cloud with 16 nodes speed up the finite element analysis over a workstation with 8 cores by more than seven-fold and eight-fold. A dedicated in-house supercomputer speeds up the finite element analysis over cloud computing by approximately two-fold because of better I/O performance and larger memory. In addition, considerable variations of elapsed time for solving the finite element model with multiple nodes in the cloud were observed due to resource sharing in cloud computing.

A virtual machine scheduler based on CPU and I/O-bound features for energy-aware in high performance computing clouds

A Virtual Machine (VM) scheduler for homogeneous High Performance Computing (HPC) cloud environments is presented in this paper. This scheduler considers each VM workload type (CPU or I/O-bound) to decide on its allocation. Scheduler is able to reduce energy consumption, as well as SLA violations on this cloud environment, avoiding performance losses by allocating simultaneously VMs which run different types of tasks.The scheduler method was validated through simulations conducted with the CloudSim framework. Two synthetic benchmarks representing both workload types were run previously in VMs in order to obtain basic data to design the scheduler. Thus, it was possible to implement a scheduling method that employs a VM allocation policy based on features of each application. Results showed that knowing the specifications and characteristics of environment may contribute to a better usage of resources, leading to an increased level of services availability and, finally, reducing problems caused by competition in resource usage.

Project Tests High-Performance Cloud Computing for Reservoir Simulations

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 167877, “Reservoir Simulations in a High-Performance Cloud-Computing Environment,” by Morgan Edward Eldred, Asma Aboubakr, Ahmed Abubakr Al-Emadi, Thomas James O’Reilly, Nedal Barghouti, and Abdollah Orangi, Maersk Oil, prepared for the 2014 SPE Intelligent Energy Conference and Exhibition, Utrecht, The Netherlands, 1–3 April. The paper has not been peer reviewed. In the upstream oil and gas industry, cloud computing is very immature because the industry has always been challenged by storage and computational capability. However, there is recent evidence for considering high-performance cloud computing (HPCC) because of the promise of benefits such as flexibility, accessibility, and cost reduction. HPCC may create an opportunity for small to midsized upstream companies that do not want to invest in the infrastructure needed for evaluating scientific applications. Project Overview The target of this project was to prove the concept of running simulation software in a high-performance computing cloud and use the findings to design a framework or methodology enabling companies to pursue business opportunities iteratively while learning along the way. The outcome of the methodology is a dynamic tactical and strategic roadmap that leverages trends in HPCC. Calculations and Results The following cases were run on a local cluster at an early stage for the purpose of run validations: HP_ICLOUD on reference workstation with one central processing unit (CPU) HP_ICLOUD on the Enterprise Cloud (ECL) server with one CPU HP_ICLOUD_4 on the ECL server with four CPUs HP_ICLOUD_8 on the ECL server with eight CPUs The four cases showed identical results for oil-production rate and cumulative oil for the duration of the field history, as expected. The case with a single CPU was completed in approximately 20 hours. The run times with four and eight CPUs were 7.7 and 5.7 hours, respectively. Fig. 1 shows that wall-clock time decreased as more CPUs were added, both for calculations performed internally and for those performed on the cloud servers. It was also observed that internal calculations stagnated at more than four CPUs (i.e., sublinear scaling). On the other hand, close to linear scaling was observed when calculations were run on the cloud servers. Assuming that this linear scaling persists when adding more than eight CPUs and extrapolating from this observation, it is hypothesized that, for larger jobs, the performance of the cloud servers would increase significantly compared with what can be achieved internally.

Evaluating and Improving the Performance and Scheduling of HPC Applications in Cloud

Cloud computing is emerging as a promising alternative to supercomputers for some high-performance computing (HPC) applications. With cloud as an additional deployment option, HPC users and providers are faced with the challenges of dealing with highly heterogeneous resources, where the variability spans across a wide range of processor configurations, interconnects, virtualization environments, and pricing models. In this paper, we take a holistic viewpoint to answer the question— why and who should choose cloud for HPC, for what applications, and how should cloud be used for HPC? To this end, we perform comprehensive performance and cost evaluation and analysis of running a set of HPC applications on a range of platforms, varying from supercomputers to clouds. Further, we improve performance of HPC applications in cloud by optimizing HPC applications’ characteristics for cloud and cloud virtualization mechanisms for HPC. Finally, we present novel heuristics for online application-aware job scheduling in multi-platform environments. Experimental results and simulations using CloudSim show that current clouds cannot substitute supercomputers but can effectively complement them. Significant improvement in average turnaround time (up to 2X)and throughput (up to 6X) can be attained using our intelligent application-aware dynamic scheduling heuristics compared tosingle-platform or application-agnostic scheduling.

A preemptive truthful VMs allocation online mechanism in private cloud

During the last decade, cloud-technology has presented considerable opportunities for high-performance computing (HPC). In addition, technical computing data centers have been able to maximize their return on investment (ROI). HPC system managers can leverage the benefits of a cloud model for their traditional HPC environments to improve scalability, simplify service access, accelerate collaboration or funding, enable pay-for-use, and improve efficiency. Many HPC clouds assume the form of private Infrastructure as a Service (IaaS). In practice, private cloud users may strategically misreport task values in order to achieve a high profit, and thus cloud providers cannot simply maximize the sum of allocated users’ value, which is called social welfare. For this reason, designing a mechanism that reveals the truthful value of users with a concern for both random arrival tasks and maximizing social welfare is necessary. In this study, a model of an online mechanism for virtual machines allocation is built, a preemptive online mechanism is proposed, the truthfulness is proved, a competitive ratio is given, and several simulations are conducted using real tasks from a data center. The total values and completed tasks are compared to the online and offline allocations, respectively, according to different capacity. The simulations reveal that our mechanism is more efficient than the offline mechanism.

Efficient network isolation and load balancing in multi-tenant HPC clusters

Multi-tenancy promises high utilization of available system resources and helps maintaining cost-effective operations for service providers. However, multi-tenant high-performance computing (HPC) infrastructures, like dynamic HPC clouds, bring unique challenges, both associated with providing performance isolation to the tenants, and achieving efficient load-balancing across the network fabric. Each tenant should experience predictable network performance, unaffected by the workload of other tenants. At the same time, it is equally important that the network links are balanced, avoiding network saturation. The network saturation can lead to unpredictable application performance, and a potential loss of profit for the cloud service providers.In this paper, we present two significant extensions to our previously proposed partition-aware fat-tree routing algorithm, pFTree, for InfiniBand-based HPC systems. First, we extend pFTree to incorporate provider defined partition-wise policies that govern how the nodes in different partitions are allowed to share network resources with each other. Second, we present a weighted version of the pFTree routing algorithm, that besides partitions, also takes node traffic characteristics into account to balance load across the network links more evenly. A comprehensive evaluation comprising both real-world experiments and simulations confirms the correctness and feasibility of the proposed extensions.

Scalability and communication performance of HPC on Azure Cloud

Different domains of research are moving to cloud computing whether to carry out compute intensive experiments or to store large datasets. Cloud computing services allow the users to focus on the application, without being concerned about the physical resources. HPC system on the cloud is desired for their high needs of efficient CPU computations. Our objective was to evaluate the scalability and performance of High Performance Cloud Computing on Microsoft Azure Cloud infrastructure by using well known Benchmarks, namely, IMB point-to-point communication and NAS Parallel Benchmarks (NPB). In our experiments, performance of the HPC applications on the cloud is assessed in terms of MOPS and speedup, and is tested under different configurations of cluster sizes. Also, point-to point communication performance between nodes is assessed in terms of latency and bandwidth as a function of message size.

Design and Implementation of Scheduling Algorithm for High Performance Cloud Computing

Cloud computing is technology that provides access to the services anywhere, anytime and paying for the services that are being used. This allows cloud users to use the data and application, if needed anywhere but must have access to the internet. The main task to be handled is the services arriving at the server for service. It is necessary to handle all services by doing proper scheduling mechanism so as to provide them service in well maintained manner. Better Quality of Service has to be provided to all clients. This strategy is used to schedule multiple user’s services which start at any time and the QoS requirements are main consideration. Two scheduling algorithms i.e. First Come First Serve and Round Robin have been executed. Experiments have been done to schedule services of fixed length as well as variable length. In both the cases this has been proved that First Come First Serve shows less average waiting time as compared to Round Robin Scheduling.

Applications for ultrascale computing

Studies of complex physical and engineering systems, represented by multi-scale and multi-physics computer simulations have an increasing demand for computing power, especially when the simulations of realistic problems are considered. This demand is driven by the increasing size and complexity of the studied systems or the time constraints. Ultrascale computing systems offer a possible solution to this problem. Future ultrascale systems will be large-scale complex computing systems combining technologies from high performance computing, distributed systems, big data, and cloud computing. Thus, the challenge of developing and programming complex algorithms on these systems is twofold. Firstly, the complex algorithms have to be either developed from scratch, or redesigned in order to yield high performance, while retaining correct functional behaviour. Secondly, ultrascale computing systems impose a number of non-functional cross-cutting concerns, such as fault tolerance or energy consumption, which can significantly impact the deployment of applications on large complex systems. This article discusses the state-of-the-art of programming for current and future large scale systems with an emphasis on complex applications. We derive a number of programming and execution support requirements by studying several computing applications that the authors are currently developing and discuss their potential and necessary upgrades for ultrascale execution.

Selected Approaches and Frameworks to Carry out Genomic Data Analysis on the Cloud

While High Performance Computing clouds allow researchers to process large amounts of genomic data, complex resource and software configuration tasks must be carried out beforehand. The current trend exposes applications and data as services, simplifying access to clouds. This paper examines commonly used cloud-based genomic analysis services, introduces the approach of exposing data as services and proposes two new solutions (HPCaaS and Uncinus) which aim to automate service development, deployment process and data provision. By comparing and contrasting these solutions, we identify key mechanisms of service creation, execution and data access required to support non-computing specialists employing clouds.

A Survey of Cloud-Based Service Computing Solutions for Mammalian Genomics

Cloud-based service computing has started to change the way how research in science, in particular biology, medicine, and engineering, is being carried out. Researchers in the area of mammalian genomics have taken advantage of cloud computing technology to cost-effectively process large amounts of data and speed up discovery. Mammalian genomics is limited by the cost and complexity of analysis, which require large amounts of computational resources to analyse huge amount of data and biology specialists to interpret results. On the other hand the application of this technology requires computing knowledge, in particular programming and operations management skills to develop high performance computing (HPC) applications and deploy them on HPC clouds. We carried out a survey of cloud-based service computing solutions, as the most recent and promising instantiations of distributed computing systems, in the context their use in research of mammalian genomic analysis. We describe our most recent research and development effort which focuses on building Software as a Service (SaaS) clouds to simplify the use of HPC clouds for carrying out mammalian genomic analysis.

Virtualizing high-end GPGPUs on ARM clusters for the next generation of high performance cloud computing

High performance cloud computing is behind the scene powering "the next big thing" as the mainstream accelerator for innovation in many areas. We describe here how to accelerate inexpensive ARM-based computing nodes with high-end GPGPUs hosted on x86_64 machines using the GVirtuS general-purpose virtualization service. We draw the vision of a possible next generation computing clusters characterized by highly heterogeneous parallelism heading to a lower electric power demanding, less heat producing and more environmental friendliness. Preliminary but promising performance data suggest that this solution could be considered as part of the foundations of the next generation of high performance cloud computing components.

Improving HPC Application Performance in Public Cloud

Improving as well as evaluating the performance of High Performance Computing (HPC) applications by migrating them to Cloud environments are widely considered as critical issues in the field of high performance and Cloud computing. However, poor network performance, heterogeneous and dynamic environments are some series of pitfalls for execution of HPC applications in Cloud. This paper proposes a new approach to improve the performance and scalability of HPC applications on Amazon's HPC Cloud. The evidence from our approach points a significant improvement in speed up and scale up with the response rate of more than 20 percent parallel efficiency on the Cloud in comparison to dedicated HPC cluster. We state that the EC2 Cloud system is a feasible platform for deploying on-demand, small sized HPC applications.

CLOUDRB: A framework for scheduling and managing High-Performance Computing (HPC) applications in science cloud

In recent years, the Cloud environment has played a major role in running High-Performance Computing (HPC) applications, which are computationally intensive and data intensive in nature. The High-Performance Computing Cloud (HPCC) or Science Cloud (SC) provides the resources to these types of applications in an on demand and scalable manner. Scheduling of jobs or applications in a Cloud environment is NP-Complete and complex in nature due to the dynamicity of resources and on demand user application requirements. The main motivation behind this research study is to design and develop a CLOUD Resource Broker (CLOUDRB) for efficiently managing cloud resources and completing jobs for scientific applications within a user-specified deadline. It is implemented and integrated with a Deadline-based Job Scheduling and Particle Swarm Optimization (PSO)-based Resource Allocation mechanism. Our proposed approach intends to achieve the objectives of minimizing both execution time and cost based on the defined fitness function. It is simulated by modeling the HPC jobs and Cloud resources using the Matlab programming environment. The simulation results prove the effectiveness of the proposed research work by minimizing the completion time, cost and job rejection ratio and maximizing the number of jobs completing their applications within a deadline and meeting the user’s satisfaction. The proposed work has been tested in our Eucalyptus-based cloud environments by submitting real-world HPC applications and observed the improvements in performance.

CCAP: A Cache Contention-Aware Virtual Machine Placement Approach for HPC Cloud

Applications in High Performance Computing (HPC) cloud are characterized by large cache resource consumption due to large-scale inputs and intensive communications, which creates serious Shared Last Level cache (SLLC) performance bottleneck. Current system software stacks are not efficient in addressing this issue among virtual machines at the hypervisor level or the threads at the operating system level. In this paper, we investigate performance interference due to contention for SLLC in the HPC cloud. We employ an enhanced reuse distance analysis technique with an accelerated cyclic compression algorithm to identify application's cache interference intensity. Based on reuse distance analysis, we propose a practical Cache Contention-Aware virtual machine Placement approach (CCAP). CCAP dispatches virtual machines according to their cache interference intensities to avoid cache pollution and interference, thus alleviating negative effects of cache contention. We implement CCAP in the Xen hypervisor. Evaluation of NPB workload reveals that CCAP can improve performance of cache sensitive applications when they are co-scheduled with cache pollution programs. For a 2-workload system, it reduces execution time by 12 %, as well as cache miss rate by 13 %, while increasing throughput by 13 %, on average. Moreover, CCAP also improves the average performance of the cache pollution programs by 5 %. For a 4-workload system, CCAP brings more significant performance improvement to cache sensitive applications, an average increase of 20 %.

A unified framework for the deployment, exposure and access of HPC applications as services in clouds

Clouds have provided on-demand, scalable and affordable High Performance Computing (HPC) resources to discipline (e.g., Biology, Medicine, Chemistry) scientists. However, the steep learning curve of preparing a HPC cloud and deploying HPC applications has hindered many scientists to achieve innovative discoveries for which HPC resources must be relied on. With the world moving to web-based tools, scientists are also seeking more web-based technologies to support their research. Unfortunately, the discipline problems of high-performance computational research are both unique and complex, which make the development of web-based tools for this research difficult. This paper presents our work on developing a unified cloud framework that allows discipline users to easily deploy and expose HPC applications in public clouds as services. To provide a proof of concept, we have implemented the cloud framework prototype by integrating three components: (i) Amazon EC2 public cloud for providing HPC infrastructure, (ii) a HPC service software library for accessing HPC resources, and (iii) the Galaxy web-based platform for exposing and accessing HPC application services. This new approach can reduce the time and money needed to deploy, expose and access discipline HPC applications in clouds.

Exploiting parallel R in the cloud with SPRINT.

Advances in DNA Microarray devices and next-generation massively parallel DNA sequencing platforms have led to an exponential growth in data availability but the arising opportunities require adequate computing resources. High Performance Computing (HPC) in the Cloud offers an affordable way of meeting this need. Bioconductor, a popular tool for high-throughput genomic data analysis, is distributed as add-on modules for the R statistical programming language but R has no native capabilities for exploiting multi-processor architectures. SPRINT is an R package that enables easy access to HPC for genomics researchers. This paper investigates: setting up and running SPRINT-enabled genomic analyses on Amazon's Elastic Compute Cloud (EC2), the advantages of submitting applications to EC2 from different parts of the world and, if resource underutilization can improve application performance. The SPRINT parallel implementations of correlation, permutation testing, partitioning around medoids and the multi-purpose papply have been benchmarked on data sets of various size on Amazon EC2. Jobs have been submitted from both the UK and Thailand to investigate monetary differences. It is possible to obtain good, scalable performance but the level of improvement is dependent upon the nature of the algorithm. Resource underutilization can further improve the time to result. End-user's location impacts on costs due to factors such as local taxation. Although not designed to satisfy HPC requirements, Amazon EC2 and cloud computing in general provides an interesting alternative and provides new possibilities for smaller organisations with limited funds.

Performance issues and performance analysis tools for HPC cloud applications: a survey

Cloud Computing is an eminent emerging technology that surpasses Grids from their IT resource administrations and arduous Grid middleware solutions. At present, users could access an abundant number of pre-defined cloud services or run their programs on demand as a pay-as-you-go processing model without much distribution problems. In addition, the IT business market has pumped enough revenue for establishing salient common-use cloud solutions. Despite adequate researchers have been involved in the cloud development, scientific application developers are still reluctant to execute their applications in the cloud due to the performance concerns, such as, scalability, availability, and service level agreement violations of the cloud providers. In this paper, a survey of various High Performance Computing (HPC) applications and possible performance concerns while executing applications in cloud is presented. Pointing out the need for Performance Analysis (PA) tools, this paper focuses on the study of cloud-based PA tools in detail. This paper could leverage HPC application developers to cope with the performance issues and to best utilize the available performance analysis tools of clouds.

INTRODUCING THE EXTREMELY HETEROGENEOUS ARCHITECTURE

The computer industry is moving towards two extremes: extremely high-performance high-throughput cloud computing, and low-power mobile computing. Cloud computing, while providing high performance, is very costly. Google and Microsoft Bing spend billions of dollars each year to maintain their server farms, mainly due to the high power bills. On the other hand, mobile computing is under a very tight energy budget, but yet the end users demand ever increasing performance on these devices. This trend indicates that conventional architectures are not able to deliver high-performance and low power consumption at the same time, and we need a new architecture model to address the needs of both extremes. In this paper, we thus introduce our Extremely Heterogeneous Architecture (EHA) project: EHA is a novel architecture that incorporates both general-purpose and specialized cores on the same chip. The general-purpose cores take care of generic control and computation. On the other hand, the specialized cores, including GPU, hard accelerators (ASIC accelerators), and soft accelerators (FPGAs), are designed for accelerating frequently used or heavy weight applications. When acceleration is not needed, the specialized cores are turned off to reduce power consumption. We demonstrate that EHA is able to improve performance through acceleration, and at the same time reduce power consumption. Since EHA is a heterogeneous architecture, it is suitable for accelerating heterogeneous workloads on the same chip. For example, data centers and clouds provide many services, including media streaming, searching, indexing, scientific computations. The ultimate goal of the EHA project is two-fold: first, to design a chip that is able to run different cloud services on it, and through this design, we would be able to greatly reduce the cost, both recurring and non-recurring, of data centers\clouds; second, to design a light-weight EHA that runs on mobile devices, providing end users with improved experience even under tight battery budget constraints.