IBM (Germany)

1. Introduction. Business Processes. Business Processes as Enterprise Resource. Virtual Enterprises. Processes and Workflows. Dimensions of Workflow. User Support. Categories of Workflows. Application Structure. Workflow and Objects. Application Operating System. Software Stack. Document/Image Processing. Groupware and Workflow. Different Views of Applications. Transactional Workflow. Advanced Usage. System Requirements. Relation to Other Technologies. 2. Business Engineering. Business Modeling. Business Logic. Enterprise Structure. Information Technology Infrastructure. Business Modeling Example. Business Process Reengineering. Process Discovery. Process Optimization. Process Analysis. Business Engineering and Workflow. Monitoring. 3. Workflow Management System Basics. Main Components. Types of Users. Buildtime. Metamodel Overview. Runtime. Audit Trail. Process Management. Authorization. Application Programming Interface. System Structure. Workflow Standards. 4. Metamodel. The Notion of a Metamodel. Process Data. Activities. Control Flow. Data Flow. Summary: PM-Graphs. Navigation. Summary: G-Instances. 5. Advanced Functions. Events. Dynamic Modification of Workflows. Advanced Join Conditions. Container Materialization. Object Staging. Context Management. Performance Spheres. Compile Spheres. 6. Workflows and Objects. Component-based Software Construction. Scripts in Object-Oriented Analysis and Design. The Object Request Broker. The OMG Workflow Management Facility. 7. Workflows and Transactions. Basic Transaction Concepts. Advanced Transaction Concepts. Streams. Atomic Spheres. Compensation Spheres. Phoenix Behavior. 8. Advanced Usage. Monitoring Dynamic Integrity Rules. Software Distribution. Security Management. Business-Process-Oriented Systems Management. 9. Application Topologies. Dependent Applications. Client/Server Structures. TP Monitors. Communication Paradigms. Message Monitors. Message Broker. Object Brokers. Distributed Applications. Web Applications. Workflow-based Applications. 10. Architecture and System Structure. Architectural Principles. System Structure. Servers. Client. Program Execution. System Group. Domains. System Tuning. Workload Management. Systems Management. Exploiting Parallel Databases. Server Implementation Aspects. Navigation. Message Queuing Usage. Process Compiler. 11. Development of Workflow-based Applications. Development Environment Blueprint. Component Generation. Testing. Animation. Debugging Activity Implementations. Application Database Design. Application Tuning. Optimization. A Travel Reservation Example. B List of Symbols. Bibliography. Index.

The final goal of all industrial machine learning (ML) projects is to develop ML products and rapidly bring them into production. However, it is highly challenging to automate and operationalize ML products and thus many ML endeavors fail to deliver on their expectations. The paradigm of Machine Learning Operations (MLOps) addresses this issue. MLOps includes several aspects, such as best practices, sets of concepts, and development culture. However, MLOps is still a vague term and its consequences for researchers and professionals are ambiguous. To address this gap, we conduct mixed-method research, including a literature review, a tool review, and expert interviews. As a result of these investigations, we contribute to the body of knowledge by providing an aggregated overview of the necessary principles, components, and roles, as well as the associated architecture and workflows. Furthermore, we provide a comprehensive definition of MLOps and highlight open challenges in the field. Finally, this work provides guidance for ML researchers and practitioners who want to automate and operate their ML products with a designated set of technologies.

This paper mainly offers guidance to the technologist who has to characterize metal‐semiconductor contacts for process development and production control. To abolish confusion clear definitions are proposed for the terms contact resistance and contact resistivity. Based on these definitions, methods of determining contact resistivity are discussed, revealing the common source of error of the methods and making comparisons. New methods like the TLM method are described and included in the comparison. Results of contact resistivity measurements on Al‐Si contacts over a wide range of silicon surface doping are presented.

Web services based on the service-oriented architecture framework provide a suitable technical foundation for making business processes accessible within enterprises and across enterprises. But to appropriately support dynamic business processes and their management, more is needed, namely, the ability to prescribe how Web services are used to implement activities within a business process, how business processes are represented as Web services, and also which business partners perform what parts of the actual business process. In this paper, the relationship between Web services and the management of business processes is worked out and presented in a tutorial-like manner.

This paper presents the results of our observing oxygen precipitation in a silicon crystal at 1000 °C using infrared absorption, TEM, and x-ray techniques. Comparison of results shows excellent agreement between TEM and IR for the number of precipitate particles formed. TEM data show that the number of particles increases as the annealing time is increased for oxygen contents above a critical concentration ratio of about 4 (30 ppma at 1000 °C). The number of particles formed at lower ratios remains relatively small. Detailed analysis of the TEM results has yielded a relationship between precipitate volume and the total area of punched-out loops. These loops, however, account for only a small fraction of the silicon atoms displaced by the growing precipitates. X-ray results indicate the defects are of an interstitial nature, in agreement with TEM.

Double-gate devices will enable the continuation of CMOS scaling after conventional scaling has stalled. DGCMOS/FinFET technology offers a tactical solution to the gate dielectric barrier and a strategic path for silicon scaling to the point where only atomic fluctuations halt further progress. The conventional nature of the processes required to fabricate these structures has enabled rapid experimental progress in just a few years. Fully integrated CMOS circuits have been demonstrated in a 180 nm foundry-compatible process, and methods for mapping conventional, planar CMOS product designs to FinFET have been developed. For both low-power and high-performance applications, DGCMOS-FinFET offers a most promising direction for continued progress in VLSI.

To meet the challenge of processing rapidly growing graph and network data created by modern applications, a number of distributed graph processing systems have emerged, such as Pregel and GraphLab. All these systems divide input graphs into partitions, and employ a "think like a vertex" programming model to support iterative graph computation. This vertex-centric model is easy to program and has been proved useful for many graph algorithms. However, this model hides the partitioning information from the users, thus prevents many algorithm-specific optimizations. This often results in longer execution time due to excessive network messages (e.g. in Pregel) or heavy scheduling overhead to ensure data consistency (e.g. in GraphLab). To address this limitation, we propose a new "think like a graph" programming paradigm. Under this graph-centric model, the partition structure is opened up to the users, and can be utilized so that communication within a partition can bypass the heavy message passing or scheduling machinery. We implemented this model in a new system, called Giraph++, based on Apache Giraph, an open source implementation of Pregel. We explore the applicability of the graph-centric model to three categories of graph algorithms, and demonstrate its flexibility and superior performance, especially on well-partitioned data. For example, on a web graph with 118 million vertices and 855 million edges, the graph-centric version of connected component detection algorithm runs 63X faster and uses 204X fewer network messages than its vertex-centric counterpart.

Oxide electronic materials provide a plethora of possible applications and offer ample opportunity for scientists to probe into some of the exciting and intriguing phenomena exhibited by oxide systems and oxide interfaces. In addition to the already diverse spectrum of properties, the nanoscale form of oxides provides a new dimension of hitherto unknown phenomena due to the increased surface-to-volume ratio.

In this paper we describe a framework for providing customers of Web services differentiated levels of service through the use of automated management and service level agreements (SLAs). The framework comprises the Web Service Level Agreement (WSLA) language, designed to specify SLAs in a flexible and individualized way, a system to provision resources based on service level objectives, a workload management system that prioritizes requests according to the associated SLAs, and a system to monitor compliance with the SLA. This framework was implemented as the utility computing services part of the IBM Emerging Technologies Tool Kit, which is publicly available on the IBM alphaWorks™ Web site.

The relevance of business processes as a major asset of an enterprise is more and more accepted: Business processes prescribe the way in which the resources of an enterprise are used, i.e., they describe how an enterprise will achieve its business goals. Organizations typically prescribe how business processes have to be performed, and they seek information technology that supports these processes. We describe a system that supports the two fundamental aspects of business process management, namely the modeling of processes and their execution. The meta-model of our system deals with models of business processes as weighted, colored, directed graphs of activities; execution is performed by navigation through the graphs according to a well-defined set of rules. The architecture consists of a distributed system with a client/server structure, and stores its data in an object-oriented database system.

This paper reviews the design challenges that current and future processors must face, with stringent power limits, high-frequency targets, and the continuing system integration trends. This paper then describes the architecture, circuit design, and physical implementation of a first-generation Cell processor and the design techniques used to overcome the above challenges. A Cell processor consists of a 64-bit Power Architecture processor coupled with multiple synergistic processors, a flexible IO interface, and a memory interface controller that supports multiple operating systems including Linux. This multi-core SoC, implemented in 90-nm SOI technology, achieved a high clock rate by maximizing custom circuit design while maintaining reasonable complexity through design modularity and reuse.

Hadoop has become an attractive platform for large-scale data analytics. In this paper, we identify a major performance bottleneck of Hadoop: its lack of ability to colocate related data on the same set of nodes. To overcome this bottleneck, we introduce CoHadoop, a lightweight extension of Hadoop that allows applications to control where data are stored. In contrast to previous approaches, CoHadoop retains the flexibility of Hadoop in that it does not require users to convert their data to a certain format (e.g., a relational database or a specific file format). Instead, applications give hints to CoHadoop that some set of files are related and may be processed jointly; CoHadoop then tries to colocate these files for improved efficiency. Our approach is designed such that the strong fault tolerance properties of Hadoop are retained. Colocation can be used to improve the efficiency of many operations, including indexing, grouping, aggregation, columnar storage, joins, and sessionization. We conducted a detailed study of joins and sessionization in the context of log processing---a common use case for Hadoop---, and propose efficient map-only algorithms that exploit colocated data partitions. In our experiments, we observed that CoHadoop outperforms both plain Hadoop and previous work. In particular, our approach not only performs better than repartition-based algorithms, but also outperforms map-only algorithms that do exploit data partitioning but not colocation. 8.

PURPOSE OF REVIEW: Electronic medical devices are an integral part of patient care. As new devices are introduced, the number of alarms to which a healthcare professional may be exposed may be as high as 1000 alarms per shift. The US Food and Drug Administration has reported over 500 alarm-related patient deaths in five years. The Joint Commission, recognizing the clinical significance of alarm fatigue, has made clinical alarm management a National Patient Safety Goal. RECENT FINDINGS: Potential solutions to alarm fatigue include technical, organizational, and educational interventions. Selecting only the right monitors (i.e., avoiding overmonitoring), judicious selection of alarm limits, and multimodal alarms can all reduce the number of nuisance alarms to which a healthcare worker is exposed. SUMMARY: Alarm fatigue can jeopardize safety, but some clinical solutions such as setting appropriate thresholds and avoiding overmonitoring are available.

In recent years, international corporations such as IBM are increasingly relying on dispersed R&D teams in order to keep pace with resource availability and the demands of global markets. The advantages of this approach arise mainly from the utilization of differences in personnel costs and gaining access to a broader knowledge base to satisfy the demands of international clients. The disadvantages of teams of this kind are obvious: geographic distances, differences in culture and work habits as well as the necessity to bridge time zones place greater demands on communication, synchronization and management. The application of specific project management methods and the intensive use of information technology (IT) lessens the disadvantages in transnational development projects. Recently completed projects involving large‐scale commercial software development at IBM demonstrate the potentials of IT in transnational development. The authors advocate the application of IT adapted to specific situations. The central fields of application of IT in dispersed R&D teams are the development of a personal network, the promotion of creativity, the exchange of technical information, and the coordination of decentralized project activities.

This is the first paper describing the IBM Blue Gene/Q interconnection network and message unit. The Blue Gene/Q system is the third generation in the IBM Blue Gene line of massively parallel supercomputers. The Blue Gene/Q architecture can be scaled to 20 PF/s and beyond. The network and the highly parallel message unit, which provides the functionality of a network interface, are integrated onto the same chip as the processors and cache memory, and consume 8% of the chip's area. For better application scalability and performance, we describe new routing algorithms and new techniques to parallelize the injection and reception of packets in the network interface. Measured hardware performance results are also presented.

Program errors detected during internal testing of the operating system DOS/VS form the basis for an investigation of error distributions in system programs. Using a classification of the errors according to various attributes, conclusions can be drawn concerning the possible causes of these errors. The information thus obtained is applied in a discussion of the most effective methods for the detection and prevention of errors.

The authors describe a novel bipolar logic featuring a direct injection of minority carriers into the switching transistor. MTL is based on inverters having decoupled multicollector outputs for the logical combinations. The devices are self-isolated and no ohmic load resistors are required. This is a key to monolithic logic chips of very high functional density and low power dissipation. On experimental chips an excellent power-delay product of 0.35 pJ has been measured. These experiments show that a density of 100 gates/mm/SUP 2/ can be achieved with present manufacturing tolerances (minimum dimensions: 0.3-mil metal line width, 0.15-mil spacing, 0.2/spl times/0.2-mil/SUP 2/ contact holes).

A significant number of companies are re-engineering their business to be more effective and productive. Consequently, existing applications must be modified, and new applications must be written. The new applications typically run in a distributed and heterogeneous environment, performing single tasks in parallel, and demanding special transaction functionality for the supporting environments. Workflow-based applications offer this type of capability. In this paper, their principal advantages are derived and set in context to transaction, object, and CASE (computer-assisted software engineering) technology. In particular, a method is proposed to develop these workflow-based applications in a cohesive and consistent way.

The IBM POWER® processor is the dominant reduced instruction set computing microprocessor in the world today, with a rich history of implementation and innovation over the last 20 years. In this paper, we describe the key features of the POWER7® processor chip. On the chip is an eight-core processor, with each core capable of four-way simultaneous multithreaded operation. Fabricated in IBM's 45-nm silicon-on-insulator (SOI) technology with 11 levels of metal, the chip contains more than one billion transistors. The processor core and caches are significantly enhanced to boost the performance of both single-threaded response-time-oriented, as well as multithreaded, throughput-oriented applications. The memory subsystem contains three levels of on-chip cache, with SOI embedded dynamic random access memory (DRAM) devices used as the last level of cache. A new memory interface using buffered double-data-rate-three DRAM and improvements in reliability, availability, and serviceability are discussed.

BlueGene/L is currently the world's fastest supercomputer. It consists of a large number of low power dual-processor compute nodes interconnected by high speed torus and collective networks, Because compute nodes do not have shared memory, MPI is the the natural programming model for this machine. The BlueGene/L MPI library is a port of MPICH2.In this paper we discuss the implementation of MPI collectives on BlueGene/L. The MPICH2 implementation of MPI collectives is based on point-to-point communication primitives. This turns out to be suboptimal for a number of reasons. Machine-optimized MPI collectives are necessary to harness the performance of BlueGene/L. We discuss these optimized MPI collectives, describing the algorithms and presenting performance results measured with targeted micro-benchmarks on real BlueGene/L hardware with up to 4096 compute nodes.