Java Bindings Research Articles

Heterogeneous Hadoop Cluster-Based Image Processing Workload Distribution Framework between CPU and GPU

Due to the rapid development of image data and the necessity to analyze it to extract meaningful information, heterogeneous systems have gained prominence. One of the most critical aspects of distributed systems is load balancing. When it comes to the distribution of workload in a balanced manner in a cluster, some heterogeneous systems are used for image processing. When workloads are allocated in these systems, the computational power of the processors is not considered. As a result, in these heterogeneous systems for image processing applications, an uneven workload distribution issue is found. A workload distribution programming framework is presented and discussed in this paper. The proposed strategy consists of two parts. As a first step, image data is split into optimal split sizes and distributed across nodes, then the image data is distributed across CPU and GPU in a second step for processing. A heterogeneous environment is created by combining the CPU and GPU. The OpenCL Java bindings are used to set up both the CPU and GPU to run the program. To assess the performance of the suggested technique, certain tests are carried out and compared to current platforms. For image processing applications in heterogeneous clusters, the proposed workload distribution approach distributed image data efficiently. The results of the proposed solution (Hadoop + GPU) show that using an effective workload allocation mechanism in heterogeneous systems reduces average execution time while improving overall application performance.

Efficient Processing of Image Processing Applications on CPU/GPU

Heterogeneous systems have gained popularity due to the rapid growth in data and the need for processing this big data to extract useful information. In recent years, many healthcare applications have been developed which use machine learning algorithms to perform tasks such as image classification, object detection, image segmentation, and instance segmentation. The increasing amount of big visual data requires images to be processed efficiently. It is common that we use heterogeneous systems for such type of applications, as processing a huge number of images on a single PC may take months of computation. In heterogeneous systems, data are distributed on different nodes in the system. However, heterogeneous systems do not distribute images based on the computing capabilities of different types of processors in the node; therefore, a slow processor may take much longer to process an image compared to a faster processor. This imbalanced workload distribution observed in heterogeneous systems for image processing applications is the main cause of inefficient execution. In this paper, an efficient workload distribution mechanism for image processing applications is introduced. The proposed approach consists of two phases. In the first phase, image data are divided into an ideal split size and distributed amongst nodes, and in the second phase, image data are further distributed between CPU and GPU according to their computation speeds. Java bindings for OpenCL are used to configure both the CPU and GPU to execute the program. The results have demonstrated that the proposed workload distribution policy efficiently distributes the images in a heterogeneous system for image processing applications and achieves 50% improvements compared to the current state-of-the-art programming frameworks.

NIX – Neuroscience information exchange format

The purpose of this document is to solicit community feedback on the Neuroscience information exchange format (NIX), a data standard for electrophysiology. This document contains the INCF standards and best practices committee's review of NIX, and the criteria in which it was evaluated (open, FAIR, testing and implementation, governance, adoption and use, stability and support, and comparison to similar standards). For the next 60 days, we are seeking community feedback on NIX. About NIX: The NIX data model allows storing fully annotated scientific datasets, i.e. the data together with rich metadata and their relations in a consistent, comprehensive format. Although developed originally for electrophysiology data, neither the data model nor the model for metadata are domain-specific to electrophysiology, but both models can be linked to predefined or custom terminologies which enables the user to give elements of the models a domain-specific, semantic context. Its aim is to achieve standardization by providing a common data structure and APIs for a multitude of data types and use cases, focused on but not limited to neuroscience. In contrast to most other approaches, the NIX approach is to achieve this flexibility with a minimum set of data model elements. The NIX project includes native I/O libraries for C++ and Python, language bindings for Java and MATLAB and a viewer for NIX data files, although the HDF5 viewer can also be used.

ICub-HRI: A Software Framework for Complex Human-Robot Interaction Scenarios on the iCub Humanoid Robot.

Generating complex, human-like behavior in a humanoid robot like the iCub requires the integration of a wide range of open source components and a scalable cognitive architecture. Hence, we present the iCub-HRI library which provides convenience wrappers for components related to perception (object recognition, agent tracking, speech recognition, and touch detection), object manipulation (basic and complex motor actions), and social interaction (speech synthesis and joint attention) exposed as a C++ library with bindings for Java (allowing to use iCub-HRI within Matlab) and Python. In addition to previously integrated components, the library allows for simple extension to new components and rapid prototyping by adapting to changes in interfaces between components. We also provide a set of modules which make use of the library, such as a high-level knowledge acquisition module and an action recognition module. The proposed architecture has been successfully employed for a complex human–robot interaction scenario involving the acquisition of language capabilities, execution of goal-oriented behavior and expression of a verbal narrative of the robot’s experience in the world. Accompanying this paper is a tutorial which allows a subset of this interaction to be reproduced. The architecture is aimed at researchers familiarizing themselves with the iCub ecosystem, as well as expert users, and we expect the library to be widely used in the iCub community.

Document classification in parallel environments using Java bindings in open MPI

This paper describes the high performance computing (HPC) of document search engines that are vectorizing each classified document. The parallelization is achieved by exploiting the parallel and distributed computing environments of collective multiple processes that are implemented by using Java message passing interface (MPI) bindings of openMPI. A parallelism model of manager/worker is implemented for obtaining the load balancing, as well as the analysis and benchmarking are achieved in our parallelism profiling model that is designed for the implementation. Two output of the experimental results are shown: the parallel processing performance with the efficiency of 80%, and the profiling results that show the utilizations and overheads in our parallelism model.

Design and implementation of Java bindings in Open MPI

This paper describes the Java MPI bindings that have been included in the Open MPI distribution. Open MPI is one of the most popular implementations of MPI, the Message-Passing Interface, which is the predominant programming paradigm for parallel applications on distributed memory computers. We have added Java support to Open MPI, exposing MPI functionality to Java programmers. Our approach is based on the Java Native Interface, and has similarities with previous efforts, as well as important differences. This paper serves as a reference for the application program interface, and in addition we provide details of the internal implementation to justify some of the design decisions. We also show some results to assess the performance of the bindings.

Integrating data storage and annotation in the data workflow using the NIX format and libraries

Integrating data storage and annotation in the data workflow using the NIX format and libraries

Integrated data acquisition, storage, retrieval and processing using the COMPASS DataBase (CDB)

We present a complex data handling system for the COMPASS tokamak, operated by IPP ASCR Prague, Czech Republic [1]. The system, called CDB (COMPASS DataBase), integrates different data sources as an assortment of data acquisition hardware and software from different vendors is used. Based on widely available open source technologies wherever possible, CDB is vendor and platform independent and it can be easily scaled and distributed. The data is directly stored and retrieved using a standard NAS (Network Attached Storage), hence independent of the particular technology; the description of the data (the metadata) is recorded in a relational database. Database structure is general and enables the inclusion of multi-dimensional data signals in multiple revisions (no data is overwritten). This design is inherently distributed as the work is off-loaded to the clients. Both NAS and database can be implemented and optimized for fast local access as well as secure remote access. CDB is implemented in Python language; bindings for Java, C/C++, IDL and Matlab are provided. Independent data acquisitions systems as well as nodes managed by FireSignal [2] are all integrated using CDB. An automated data post-processing server is a part of CDB. Based on dependency rules, the server executes, in parallel if possible, prescribed post-processing tasks.

FastMPJ: a scalable and efficient Java message-passing library

The performance and scalability of communications are key for high performance computing (HPC) applications in the current multi-core era. Despite the significant benefits (e.g., productivity, portability, multithreading) of Java for parallel programming, its poor communications support has hindered its adoption in the HPC community. This paper presents FastMPJ, an efficient message-passing in Java (MPJ) library, boosting Java for HPC by: (1) providing high-performance shared memory communications using Java threads; (2) taking full advantage of high-speed cluster networks (e.g., InfiniBand) to provide low-latency and high bandwidth communications; (3) including a scalable collective library with topology aware primitives, automatically selected at runtime; (4) avoiding Java data buffering overheads through zero-copy protocols; and (5) implementing the most widely extended MPI-like Java bindings for a highly productive development. The comprehensive performance evaluation on representative testbeds (InfiniBand, 10 Gigabit Ethernet, Myrinet, and shared memory systems) has shown that FastMPJ communication primitives rival native MPI implementations, significantly improving the efficiency and scalability of Java HPC parallel applications.

The Next Generation ARC Middleware and ATLAS Computing Model

The distributed NDGF Tier-1 and associated NorduGrid clusters are well integrated into the ATLAS computing environment but follow a slightly different paradigm than other ATLAS resources. The current paradigm does not divide the sites as in the commonly used hierarchical model, but rather treats them as a single storage endpoint and a pool of distributed computing nodes. The next generation ARC middleware with its several new technologies provides new possibilities in development of the ATLAS computing model, such as pilot jobs with pre-cached input files, automatic job migration between the sites, integration of remote sites without connected storage elements, and automatic brokering for jobs with non-standard resource requirements. ARC's data transfer model provides an automatic way for the computing sites to participate in ATLAS’ global task management system without requiring centralised brokering or data transfer services. The powerful API combined with Python and Java bindings can easily be used to build new services for job control and data transfer. Integration of the ARC core into the EMI middleware provides a natural way to implement the new services using the ARC components

Data Parallel Skeletons in Java

In the past years, multi-core processors and clusters of multi-core processors have emerged to be promising approaches to meet the growing demand for computing performance. They deliver scalable performance, certainly at the costs of tedious and complex parallel programming. Due to a lack of high-level abstractions, developers of parallel applications have to deal with low-level details such as coordinating threads or synchronizing processes. Thus, parallel programming still remains a dificult and error-prone task. In order to shield the programmer from these low–level details, algorithmic skeletons have been proposed. They encapsulate typical parallel programming patterns and have emerged to be an effcient and scalable approach to simplifying the development of parallel applications. In this paper, we present a Java binding of our skeleton library Muesli. We point out strengths and weaknesses of Java with respect to parallel and distributed computing. A matrix multiplication benchmark demonstrates that the Java Generics deliver poor performance, thus the Java implementation is unable to compete with the C++ implementation in terms of performance.

Implementation of the Lattice Boltzmann Method on Heterogeneous Hardware and Platforms using OpenCL

The Lattice Boltzmann method (LBM) has become an alternative method for computational fluid dynamics with a wide range of applications. Besides its numerical stability and accuracy, one of the major advantages of LBM is its relatively easy parallelization and, hence, it is especially well fitted to many-core hardware as graphics processing units (GPU). The majority of work concerning LBM implementation on GPU’s has used the CUDA programming model, supported exclusively by NVIDIA. Recently, the open standard for parallel programming of heterogeneous systems (OpenCL) has been introduced. OpenCL standard matures and is supported on processors from most vendors. In this paper, we make use of the OpenCL framework for the lattice Boltzmann method simulation, using hardware accelerators – AMD ATI Radeon GPU, AMD Dual-Core CPU and NVIDIA GeForce GPU’s. Application has been developed using a combination of Java and OpenCL programming languages. Java bindings for OpenCL have been utilized. This approach offers the benefits of hardware and operating system independence, as well as speeding up of lattice Boltzmann algorithm. It has been showed that the developed lattice Boltzmann source code can be executed without modification on all of the used hardware accelerators. Performance results have been presented and compared for the hardware accelerators that have been utilized.

JSBML: a flexible and entirely Java-based library for working with SBML

AbstractThe XML-based Systems Biology Markup Language (SBML) is the de facto standard file format for the storage and exchange of biochemical network models, and is supported by more than 180 software packages to date. Much of this success is due to its clearly defined specifications and the availability of libSBML, a portable, robust, and easy-to-use library.LibSBML provides many methods for manipulating and validating SBML files through its Application Programming Interface (API). Originally written in C and C++, libSBML also provides automatically-generated language bindings for Java, MATLAB, Perl, and many more. However, the platform independence of Java is compromised in libSBML due to the fact that the language binding is a wrapper around the C/C++ core. The real-world experiences of many software developers has been that the deployment of portable libSBML-based Java applications is difficult. Further, the libSBML API and type hierarchy are not sufficiently intuitive from a Java programmer's perspective just because they were not designed directly for Java. For these reasons, several groups in the SBML community have mounted an open-source effort to develop a pure Java library for SBML. Here we present the JSBML project, whose products are freely available at the web site http://sourceforge.net/projects/jsbml.

JSBML

AbstractThe XML-based Systems Biology Markup Language (SBML) is the de facto standard file format for the storage and exchange of biochemical network models, and is supported by more than 180 software packages to date. Much of this success is due to SBML's clearly defined specifications and the availability of libSBML, a portable, robust and easy-to-use library.LibSBML provides many methods for manipulating and validating SBML files through its Application Programming Interface (API). Originally written in C and C++, libSBML also provides automatically-generated language bindings for Java, MATLAB, Perl, Python, and many more. However, the platform independence of Java is compromised in libSBML due to the fact that the language binding is a wrapper around the C/C++ core. The real-world experiences of many software developers has been that the deployment of portable libSBML-based Java applications is difficult. Further, the libSBML API and type hierarchy are not sufficiently intuitive from a Java programmer's perspective simply because they were not designed directly for Java.For these reasons, several groups in the SBML community have mounted an open-source effort to develop a pure Java library for SBML. Here we present the JSBML project, whose products are freely available at the website "http://jsbml.sourceforge.net/":http://jsbml.sourceforge.net/. The JSBML project's aim is to provide an SBML parser and programming library that maps all SBML elements to a flexible and extended type hierarchy. Where possible, JSBML strives to attain 100% API compatibility with the libSBML Java API, to facilitate a switch from one library to the other for those who desire it.Currently, JSBML supports all constructs for SBML up to the latest Level 3 Version 1 release candidate specification, including an API to add SBML extensions. There are no plans to re-implement some of the more complex functions of libSBML, since separate community efforts are expected to make them available to JSBML via web services. These more complex functions include model consistency checking, SBML validation, and the conversion between different SBML Levels and Versions.In summary, JSBML is a young, ongoing software project that provides comprehensive and entirely Java-based data structures to read, write, and manipulate SBML files. Its layered architecture allows for the creation of Java web start applications and CellDesigner plug-ins based on stand-alone programs with very little effort. One program, SBMLsqueezer 1.3, has already been re-implemented and released using JSBML, and many others are planned.

JSBML

The XML-based Systems Biology Markup Language (SBML) is the de facto standard file format for the storage and exchange of biochemical network models, and is supported by more than 180 software packages to date. Much of this success is due to SBML's clearly defined specifications and the availability of libSBML, a portable, robust and easy-to-use library.LibSBML provides many methods for manipulating and validating SBML files through its Application Programming Interface (API). Originally written in C and C++, libSBML also provides automatically-generated language bindings for Java, MATLAB, Perl, Python, and many more. However, the platform independence of Java is compromised in libSBML due to the fact that the language binding is a wrapper around the C/C++ core. The real-world experiences of many software developers has been that the deployment of portable libSBML-based Java applications is difficult. Further, the libSBML API and type hierarchy are not sufficiently intuitive from a Java programmer's perspective simply because they were not designed directly for Java.For these reasons, several groups in the SBML community have mounted an open-source effort to develop a pure Java library for SBML. Here we present the JSBML project, whose products are freely available at the website "http://jsbml.sourceforge.net/":http://jsbml.sourceforge.net/. The JSBML project's aim is to provide an SBML parser and programming library that maps all SBML elements to a flexible and extended type hierarchy. Where possible, JSBML strives to attain 100% API compatibility with the libSBML Java API, to facilitate a switch from one library to the other for those who desire it.Currently, JSBML supports all constructs for SBML up to the latest Level 3 Version 1 release candidate specification, including an API to add SBML extensions. There are no plans to re-implement some of the more complex functions of libSBML, since separate community efforts are expected to make them available to JSBML via web services. These more complex functions include model consistency checking, SBML validation, and the conversion between different SBML Levels and Versions.In summary, JSBML is a young, ongoing software project that provides comprehensive and entirely Java-based data structures to read, write, and manipulate SBML files. Its layered architecture allows for the creation of Java web start applications and CellDesigner plug-ins based on stand-alone programs with very little effort. One program, SBMLsqueezer 1.3, has already been re-implemented and released using JSBML, and many others are planned.

Visual Workshop as the Single Model Editor for the ANSWERS Shielding and Criticality Codes

In this paper, we describe the design philosophy behind Visual Workshop, a model-viewing program. Visual Workshop is being developed as part of the nuclear codes development consortium between BNFL Group and Serco Assurance and will be marketed via Serco Assurance’s ANSWERS Software Service. Visual Workshop is a Java-based program that allows shielding and criticality engineers to visualize their computational models for the ANSWERS computational codes in a multiplicity of ways—wireframe and two-dimensional and three-dimensional ray trace—via a single graphical interface.Visual Workshop integrates the cross-platform portability of Java with the existing FORTRAN calculation engines that are used successfully with the existing ANSWERS visualization codes. The use of the Swing graphical user interface library provides a common look-and-feel on all platforms. Visual Workshop also has the innovation of employing JOGL, the Java binding to OpenGL, thus gaining the advantages of hardware support that OpenGL can bring. Visual Workshop has also made use of open-source freely licensed applications to provide functionality and to speed up development.Visual Workshop will act as a single interface for the checking of models, the submission of jobs, and the examination of job output. As such it builds upon the experiences gained using the suite of ANSWERS tools by expanding the existing capabilities to allow for the selective display of parameterized models, allow more effective error checking, and provide additional postprocessing and results display capabilities.Visual Workshop provides capabilities and tools for the engineers to refine and verify their models in a way that enhances their productivity.

LibSBML: an API Library for SBML

LibSBML is an application programming interface library for reading, writing, manipulating and validating content expressed in the Systems Biology Markup Language (SBML) format. It is written in ISO C and C++, provides language bindings for Common Lisp, Java, Python, Perl, MATLAB and Octave, and includes many features that facilitate adoption and use of both SBML and the library. Developers can embed libSBML in their applications, saving themselves the work of implementing their own SBML parsing, manipulation and validation software. LibSBML 3 was released in August 2007. Source code, binaries and documentation are freely available under LGPL open-source terms from http://sbml.org/software/libsbml.

Genericity in Java: persistent and database systems implications

Lack of parametric polymorphism has been a major obstacle for making Java a viable database programming language. Regrettably, a recently accepted solution for genericity in Java 5.0 has far-reaching negative implications for persistent and database systems because of static and dynamic type violations. Severe implications occur in typical database transactions when processing a variety of database collections. Well-known approaches to persistence in Java, including Java's own persistence mechanism, do not perform correctly due to incorrect dynamic type information that gets promoted to persistence along with objects. Dynamic checking of types of objects fetched from the persistent store may now lead to unexpected type violations. Further problems occur in reflective transactions as Java Core Reflection now allows dynamic type violations without detecting them or throwing standard exceptions. All of this shows that extending Java with parametric polymorphism has not made Java a more viable database programming language. Both legacy systems, such as those based on the Java binding of the ODMG or JDO, and future Java-related persistent and database technologies will be affected. The source of these problems is in an implementation idiom called type erasure. This paper provides formal proofs of the above implications of type erasure along with specific samples of code in Java 5.0 illustrating these violations. The limitations of the virtual platform and extensions required for persistent systems to solve this problem are also elaborated.

A Hierarchical 3D Data Rendering System Synchronized with HTML

We propose a new rendering system for large-scale, 3D geometic data that can be used with web-based content man-agement systems (CMS). To achieve this, we employed a geometry hierarchical encoding method "QSplat" and implemented this in a Java and JOGL (Java bindings of OpenGL) environment. Users can view large-scale geometric data using conventional HTML browsers with a non-powerful CPU and low-speed networks. Further, this system is independent of the platforms. We add new functionalities so that users can easily understand the geometric data: Annotations and HTML Synchronization. Users can see the geometric data with the associated annotations that describe the names or the detailed explanations of the particular portions. The HTML Synchronization enables users to smoothly and interactively switch our rendering system and HTML contents. The experimental results show that our system performs an interactive frame rate even for a large-scale data whereas other systems cannot render them

Web-Based Interactive 3D Visualization for Computer Graphics Education

Instead of developing web-based course material using an off-the-shelf web authoring tool, we believe that a better way to show principles and techniques in computer science is to have the related algorithms running “live” in the background and to allow students to interact with them within a web browser. We have chosen computer graphics as an example course because of its demand for visualization and its technical challenge for 3D rendering. This paper presents a set of web-based demos designed to demonstrate computer graphics concepts and OpenGL functions. We have ported to the web the popular Nate Robins’ OpenGL demos. We have also developed our own demos to address pedagogical aspects of computer graphics education. The main idea of the approach is to put a real-world scene and a rendered result side by side together with a set of OpenGL functions to produce the rendered result from the real-world scene. Animation, user interaction, manipulation, and virtual navigation are supported in the sense that the order of the functions and the parameters of each function can be changed interactively and such changes will be reflected immediately in the rendered result. The demos are written by using Java and GL4Java, an OpenGL Java binding to ensure the deployment on the Web. The result is a set of web-based interactive tutorials with rich visualization driven by underlying algorithms to demonstrate the subject principles and techniques. Although the contents of these demos are specific to computer graphics, the presented methodology represents a generic approach which is discipline/course independent and can be applied to various other courses.