computersystems have achieved significant progress in the areas of technology, performance, capability, and RAS (reliability/availability/serviceability) during the last quarter century. In this paper, we shall revie...
详细信息
computersystems have achieved significant progress in the areas of technology, performance, capability, and RAS (reliability/availability/serviceability) during the last quarter century. In this paper, we shall review the advances of IBM computersystems in the RAS area. This progress has for the most part been evolutionary; however, in some cases it has been revolutionary. RAS developments have been driven primarily by technological advances and by increases in functional capability and complexity, but RAS considerations have also played a leading role and have improved technological and functional capability. The paper briefly reviews the progress of computer technology. It points out how IBM has maintained or improved its systems RAS capabilities in the face of the greatly increased number of components and system complexity by improved system recovery and serviceability capability, as well as by basic improvements in intrinsic component failure rate. The paper also covers the CPU, tape, and disk areas and shows how RAS improvements in these areas have been significant. The main objective is to provide a comprehensive view of significant developments in the RAS characteristics of IBM computersystems over the past twenty-five years.
Programming of an adaptive information security system (ISS) is reduced to describing of information fields of neural networks (NN) in the form of batch neural-network programs. A similar description of data-field com...
详细信息
Programming of an adaptive information security system (ISS) is reduced to describing of information fields of neural networks (NN) in the form of batch neural-network programs. A similar description of data-field component makes it possible to study processes of operation and adaptation of neural-network ISS in the composition of computer-aided systems by modeling the interaction of operating data with the information field of NN.
Techniques of statistical design of experiments have been successfully employed for many decades in a variety of applications in industry, agriculture, medicine, psychology, and other physical and social sciences. The...
详细信息
Techniques of statistical design of experiments have been successfully employed for many decades in a variety of applications in industry, agriculture, medicine, psychology, and other physical and social sciences. Their aim is to provide scientific and efficient means of studying the effects, on one or more variables of interest, of varying multiple controllable factors in an experiment. These techniques have not been widely used in the study of computersystems, although they can potentially have as large an impact as they have had in other fields. The purpose of this paper is to review some of the basic concepts underlying the statistical design and analysis of experiments and to illustrate them by means of examples drawn from studies of computer system performance. The examples include comparisons of alternate page replacement and free storage management algorithms, optimization of a scheduler, and validation of a system simulation model.
Performance modeling can be used throughout the life of a computer system, from initial design, through implementation, configuration (and reconfiguration) and even tuning. Performance models are usually solved by num...
详细信息
Performance modeling can be used throughout the life of a computer system, from initial design, through implementation, configuration (and reconfiguration) and even tuning. Performance models are usually solved by numerical techniques, where possible, and by simulation, otherwise. This paper summarizes IBM's contributions to performance modeling and the solution of performance models.
The authors discuss the static paralleling of an initial program during the source of translation, which involves the following: (1) the paralleling of loops (the transformation of scalar operations inside a loop into...
详细信息
The authors discuss the static paralleling of an initial program during the source of translation, which involves the following: (1) the paralleling of loops (the transformation of scalar operations inside a loop into vector operations);(2) the paralleling of linear segments (the organization of conveyer-parallel execution of instructions);and (3) the detection of parallelisms between fragments of the program and formation of parallel branches. The former two aspects constitute the local paralleling, while the third aspect is the global paralleling. They then consider the detection of parallelisms between individual program fragments and briefly discuss program structuring.
This paper discusses the design and building of an APL interpreter for the IBM Personal computer. Discussed is the writing of the interpreter itself, which required the use of an intermediate language designed by the ...
详细信息
This paper discusses the design and building of an APL interpreter for the IBM Personal computer. Discussed is the writing of the interpreter itself, which required the use of an intermediate language designed by the authors. This machine-independent language also made possible the development of APL interpreters for two other systems—stem/370 and Series/1. The particularizing of the interpreter required a compiler, which in the case of the Personal computer produced Intel 8088 and 8087 assembly language code. The matching of the APL interpreter to the operating system (DOS) required an APL supervisor, which is also discussed in this paper. The provision of the APL character set presented problems, the solutions of which are also presented. Other topics discussed are the display, the keyboard, and the session manager.
Consideration was given to the key definitions, notions, and models that may be useful for transient-fault-tolerant and fault-tolerant computing in the unmanned multimachine computersystems having many interconnected...
详细信息
Consideration was given to the key definitions, notions, and models that may be useful for transient-fault-tolerant and fault-tolerant computing in the unmanned multimachine computersystems having many interconnected autonomous computers without shared memory and centralized control organ and operating with high degree of computational parallelism, that is, executing on different computers simultaneously various tasks which interchange their information. These computations should establish reliable results under byzantine faults and controllable degradation of the system at detection of faults.
This paper discusses the need for a systematic approach to computersystems design whereby networks of files and programs can be planned in a logical, orderly manner. Alternative methods are compared and developed int...
详细信息
This paper discusses the need for a systematic approach to computersystems design whereby networks of files and programs can be planned in a logical, orderly manner. Alternative methods are compared and developed into a design ‘ladder’; each step in this ladder represents a design decision and an eventual climb of the ladder achieves a feasible computer system.
Described here is an electromagnetic approach for the analysis of high-performance computer packages such as the thermal conduction module (TCM) used in the IBM 3080 and 3090 processor units. Modeling of signal paths ...
详细信息
Described here is an electromagnetic approach for the analysis of high-performance computer packages such as the thermal conduction module (TCM) used in the IBM 3080 and 3090 processor units. Modeling of signal paths and limitations of previous methods are discussed. Numerical results are presented for propagation characteristics associated with signal lines and vias, and for coupled noise between signal lines. The results are compared with those obtained by means of test vehicles, scale models, and capacitance calculations.
This paper investigates a stochastic model related to a xenix operatingcomputer system. In this system one console and two dumb terminals are arranged in parallel and connected with a central processing unit (CPU). A...
详细信息
This paper investigates a stochastic model related to a xenix operatingcomputer system. In this system one console and two dumb terminals are arranged in parallel and connected with a central processing unit (CPU). A dumb terminal may fail due to hardware while a console may fail due to hardware or due to software. Two repair facilities are available, one for inspection and repair of a console terminal and the other for the repair of a dumb terminal. Using the regenerative point technique, various reliability measures of system effectiveness are obtained.
暂无评论