scholarly journals High Performance Gibbs Sampling for IRT Models Using Row-Wise Decomposition

2012 ◽  
Vol 2012 ◽  
pp. 1-9 ◽  
Author(s):  
Yanyan Sheng ◽  
Mona Rahimi
Keyword(s):  
High Performance ◽  
Communication Overhead ◽  
Data Dependencies ◽  
Decomposition Scheme ◽  
High Data ◽  
Irt Models ◽  
Multiple Processors ◽  
The Cost ◽  
Fully Bayesian ◽  
Performance Computing

Item response theory (IRT) is a popular approach used for addressing statistical problems in psychometrics as well as in other fields. The fully Bayesian approach for estimating IRT models is computationally expensive. This limits the use of the procedure in real applications. In an effort to reduce the execution time, a previous study shows that high performance computing provides a solution by achieving a considerable speedup via the use of multiple processors. Given the high data dependencies in a single Markov chain for IRT models, it is not possible to avoid communication overhead among processors. This study is to reduce communication overhead via the use of a row-wise decomposition scheme. The results suggest that the proposed approach increased the speedup and the efficiency for each implementation while minimizing the cost and the total overhead. This further sheds light on developing high performance Gibbs samplers for more complicated IRT models.

scholarly journals A GPU-Based Gibbs Sampler for a Unidimensional IRT Model

2014 ◽  
Vol 2014 ◽  
pp. 1-11
Author(s):  
Yanyan Sheng ◽  
William S. Welling ◽  
Michelle M. Zhu
Keyword(s):  
Message Passing ◽  
Large Scale ◽  
Message Passing Interface ◽  
Cost Effective ◽  
Communication Overhead ◽  
Graphic Processing Units ◽  
Data Dependencies ◽  
High Data ◽  
Irt Models ◽  
Fully Bayesian

Item response theory (IRT) is a popular approach used for addressing large-scale statistical problems in psychometrics as well as in other fields. The fully Bayesian approach for estimating IRT models is usually memory and computationally expensive due to the large number of iterations. This limits the use of the procedure in many applications. In an effort to overcome such restraint, previous studies focused on utilizing the message passing interface (MPI) in a distributed memory-based Linux cluster to achieve certain speedups. However, given the high data dependencies in a single Markov chain for IRT models, the communication overhead rapidly grows as the number of cluster nodes increases. This makes it difficult to further improve the performance under such a parallel framework. This study aims to tackle the problem using massive core-based graphic processing units (GPU), which is practical, cost-effective, and convenient in actual applications. The performance comparisons among serial CPU, MPI, and compute unified device architecture (CUDA) programs demonstrate that the CUDA GPU approach has many advantages over the CPU-based approach and therefore is preferred.

Constructing a Bioinformatics Platform with Web and Mobile Services Based on NVIDIA Jetson TK1

2015 ◽  
Vol 7 (4) ◽  
pp. 57-73 ◽  
Author(s):  
Chun-Yuan Lin ◽  
Jin Ye ◽  
Che-Lun Hung ◽  
Chung-Hung Wang ◽  
Min Su ◽  
...  
Keyword(s):  
Power Consumption ◽  
High Performance Computing ◽  
Graphics Processing Units ◽  
High Performance ◽  
Low Cost ◽  
Research Direction ◽  
Mobile Services ◽  
Performance Ratio ◽  
The Cost ◽  
Performance Computing

Current high-end graphics processing units (abbreviate to GPUs), such as NVIDIA Tesla, Fermi, Kepler series cards which contain up to thousand cores per-chip, are widely used in the high performance computing fields. These GPU cards (called desktop GPUs) should be installed in personal computers/servers with desktop CPUs; moreover, the cost and power consumption of constructing a high performance computing platform with these desktop CPUs and GPUs are high. NVIDIA releases Tegra K1, called Jetson TK1, which contains 4 ARM Cortex-A15 CPUs and 192 CUDA cores (Kepler GPU) and is an embedded board with low cost, low power consumption and high applicability advantages for embedded applications. NVIDIA Jetson TK1 becomes a new research direction. Hence, in this paper, a bioinformatics platform was constructed based on NVIDIA Jetson TK1. ClustalWtk and MCCtk tools for sequence alignment and compound comparison were designed on this platform, respectively. Moreover, the web and mobile services for these two tools with user friendly interfaces also were provided. The experimental results showed that the cost-performance ratio by NVIDIA Jetson TK1 is higher than that by Intel XEON E5-2650 CPU and NVIDIA Tesla K20m GPU card.

Advanced Simulation Capability for Environmental Management: Current Status and Future Applications

2013 ◽  
Author(s):  
Mark Freshley ◽  
Paul Dixon ◽  
Paul Black ◽  
Bruce Robinson ◽  
Tom Stockton ◽  
...  
Keyword(s):  
Environmental Management ◽  
Radioactive Waste ◽  
Decision Analysis ◽  
High Performance Computing ◽  
Waste Disposal ◽  
High Performance ◽  
Fate And Transport ◽  
Radioactive Waste Disposal ◽  
The Cost ◽  
Performance Computing

The U.S. Department of Energy (USDOE) Office of Environmental Management (EM), Office of Soil and Groundwater (EM-12), is supporting development of the Advanced Simulation Capability for Environmental Management (ASCEM). ASCEM is a state-of-the-art scientific tool and approach that is currently aimed at understanding and predicting contaminant fate and transport in natural and engineered systems. ASCEM is a modular and open source high-performance computing tool. It will be used to facilitate integrated approaches to modeling and site characterization, and provide robust and standardized assessments of performance and risk for EM cleanup and closure activities. The ASCEM project continues to make significant progress in development of capabilities, with current emphasis on integration of capabilities in FY12. Capability development is occurring for both the Platform and Integrated Toolsets and High-Performance Computing (HPC) multiprocess simulator. The Platform capabilities provide the user interface and tools for end-to-end model development, starting with definition of the conceptual model, management of data for model input, model calibration and uncertainty analysis, and processing of model output, including visualization. The HPC capabilities target increased functionality of process model representations, toolsets for interaction with Platform, and verification and model confidence testing. The integration of the Platform and HPC capabilities were tested and evaluated for EM applications in a set of demonstrations as part of Site Applications Thrust Area activities in 2012. The current maturity of the ASCEM computational and analysis capabilities has afforded the opportunity for collaborative efforts to develop decision analysis tools to support and optimize radioactive waste disposal. Recent advances in computerized decision analysis frameworks provide the perfect opportunity to bring this capability into ASCEM. This will allow radioactive waste disposal to be evaluated based on decision needs, such as disposal, closure, and maintenance. Decision models will be used in ASCEM to identify information/data needs, and model refinements that might be necessary to effectively reduce uncertainty in waste disposal decisions. Decision analysis models start with tools for framing the problem, and continue with modeling both the science side of the problem (for example, inventories, source terms, fate and transport, receptors, risk, etc.), and the cost side of the problem, which could include costs of implementation of any action that is chosen (e.g., for disposal or closure), and the values associated with those actions. The cost side of the decision problem covers economic, environmental and societal costs, which correspond to the three pillars of sustainability (economic, social, and environmental). These tools will facilitate stakeholder driven decision analysis to support optimal sustainable solutions in ASCEM.

scholarly journals Trends in high-performance computing for engineering calculations

2014 ◽  
Vol 372 (2022) ◽  
pp. 20130319 ◽  
Author(s):  
M. B. Giles ◽  
I. Reguly
Keyword(s):  
High Performance Computing ◽  
High Performance ◽  
Full Potential ◽  
Hardware Complexity ◽  
The Past ◽  
Key Issues ◽  
The Cost ◽  
Application Developers ◽  
Near Future ◽  
Performance Computing

High-performance computing has evolved remarkably over the past 20 years, and that progress is likely to continue. However, in recent years, this progress has been achieved through greatly increased hardware complexity with the rise of multicore and manycore processors, and this is affecting the ability of application developers to achieve the full potential of these systems. This article outlines the key developments on the hardware side, both in the recent past and in the near future, with a focus on two key issues: energy efficiency and the cost of moving data. It then discusses the much slower evolution of system software, and the implications of all of this for application developers.

Abstractions and Middleware for Petascale Computing and Beyond

2012 ◽  
pp. 161-178
Author(s):  
Ivo F. Sbalzarini
Keyword(s):  
Numerical Simulations ◽  
High Performance Computing ◽  
Data Structures ◽  
High Performance ◽  
Source Code ◽  
Limiting Factors ◽  
Communication Overhead ◽  
Parallel Data ◽  
Petascale Computing ◽  
Performance Computing

As high-performance computing moves to the petascale and beyond, a number of algorithmic and software challenges need to be addressed. This paper reviews the main performance-limiting factors in today’s high-performance computing software and outlines a possible new programming paradigm to address them. The proposed paradigm is based on abstract parallel data structures and operations that encapsulate much of the complexity of an application, but still make communication overhead explicit. The authors argue that all numerical simulations can be formulated in terms of the presented abstractions, which thus define an abstract semantic specification language for parallel numerical simulations. Simulations defined in this language can automatically be translated to source code containing the appropriate calls to a middleware that implements the underlying abstractions. Finally, the structure and functionality of such a middleware are outlined while demonstrating its feasibility on the example of the parallel particle-mesh library (PPM).

Scheduling Data Intensive Scientific Workflows in Cloud Environment Using Nature Inspired Algorithms

2019 ◽  
pp. 196-217 ◽  
Author(s):  
Shikha Mehta ◽  
Parmeet Kaur
Keyword(s):  
High Performance ◽  
Heuristic Algorithms ◽  
Control Flow ◽  
Grey Wolf ◽  
Data Intensive ◽  
Np Hard Problem ◽  
Shuffled Frog Leaping ◽  
Nature Inspired Algorithms ◽  
The Cost ◽  
Performance Computing

Workflows are a commonly used model to describe applications consisting of computational tasks with data or control flow dependencies. They are used in domains of bioinformatics, astronomy, physics, etc., for data-driven scientific applications. Execution of data-intensive workflow applications in a reasonable amount of time demands a high-performance computing environment. Cloud computing is a way of purchasing computing resources on demand through virtualization technologies. It provides the infrastructure to build and run workflow applications, which is called ‘Infrastructure as a Service.' However, it is necessary to schedule workflows on cloud in a way that reduces the cost of leasing resources. Scheduling tasks on resources is a NP hard problem and using meta-heuristic algorithms is an obvious choice for the same. This chapter presents application of nature-inspired algorithms: particle swarm optimization, shuffled frog leaping algorithm and grey wolf optimization algorithm to the workflow scheduling problem on the cloud. Simulation results prove the efficacy of the suggested algorithms.

Abstractions and Middleware for Petascale Computing and Beyond

2010 ◽  
Vol 1 (2) ◽  
pp. 40-56 ◽  
Author(s):  
Ivo F. Sbalzarini
Keyword(s):  
Numerical Simulations ◽  
High Performance Computing ◽  
High Performance ◽  
Source Code ◽  
Limiting Factors ◽  
Communication Overhead ◽  
Programming Paradigm ◽  
Parallel Data ◽  
Petascale Computing ◽  
Performance Computing

As high-performance computing moves to the petascale and beyond, a number of algorithmic and software challenges need to be addressed. This paper reviews the main performance-limiting factors in today’s high-performance computing software and outlines a possible new programming paradigm to address them. The proposed paradigm is based on abstract parallel data structures and operations that encapsulate much of the complexity of an application, but still make communication overhead explicit. The authors argue that all numerical simulations can be formulated in terms of the presented abstractions, which thus define an abstract semantic specification language for parallel numerical simulations. Simulations defined in this language can automatically be translated to source code containing the appropriate calls to a middleware that implements the underlying abstractions. Finally, the structure and functionality of such a middleware are outlined while demonstrating its feasibility on the example of the parallel particle-mesh library (PPM).

scholarly journals A Comparative Study: Taxonomy of High Performance Computing (HPC)

2018 ◽  
Vol 8 (5) ◽  
pp. 3386
Author(s):  
Ranjit Rajak
Keyword(s):  
Comparative Study ◽  
High Performance Computing ◽  
High Performance ◽  
Cluster Computing ◽  
Market Demand ◽  
Huge Data ◽  
Computing Grid ◽  
Software And Hardware ◽  
The Cost ◽  
Performance Computing

The computer technologies have rapidly developed in both software and hardware field. The complexity of software is increasing as per the market demand because the manual systems are going to become automation as well as the cost of hardware is decreasing. High Performance Computing (HPC) is very demanding technology and an attractive area of computing due to huge data processing in many applications of computing. The paper focus upon different applications of HPC and the types of HPC such as Cluster Computing, Grid Computing and Cloud Computing. It also studies, different classifications and applications of above types of HPC. All these types of HPC are demanding area of computer science. This paper also done comparative study of grid, cloud and cluster computing based on benefits, drawbacks, key areas of research, characterstics, issues and challenges.

Stronger trust and privacy in social networks via local cooperation1

2019 ◽  
Vol 8 (4) ◽  
Author(s):  
Krzysztof Grining ◽  
Marek Klonowski ◽  
Malgorzata Sulkowska
Keyword(s):  
Social Networks ◽  
Real World ◽  
High Performance ◽  
Preferential Attachment ◽  
Formal Analysis ◽  
Real Life ◽  
Communication Overhead ◽  
Power Law Distribution ◽  
Large Groups ◽  
Performance Computing

Abstract In our article, we present several protocols that allow to efficiently construct large groups of users based only on local relations of trust. What is more, our approach proves to need only very small computational and communication overhead. Moreover, we give guarantees that a trusted core of the network is defended, even facing a powerful adversary capable of controlling a vast majority of users. This is non-trivial property in real-life networks, as those are usually modelled using preferential attachment graphs, which are extremely prone to attacks on the hub nodes. We show that using our protocols we can achieve similar robustness as Erdős–Renyí graphs, which, on the contrary, are very resistant against attacks focused on chosen nodes. Our protocols have been tested on graphs representing real-world social networks using high performance computing due to the size of the networks. In addition for some protocols, we provided a formal analysis to prove some phenomena in random graphs following power-law distribution, which we use as a network model. Finally, we explicitly demonstrate how our approach can be used to amplify security offered by some privacy-preserving protocols. We believe however that our results can be also seen as a contribution to fundamental observation about the nature of social networks. These results may help to design protocols, whenever it is necessary to gather a big group of users in highly dynamic or even adversarial settings.

Sign in / Sign up

Export Citation Format

Share Document