Showing papers on "Electronic design automation published in 1978"

SCALD: Structured Computer-Aided Logic Design

Abstract: SCALD, a graphics-based hierarchical digital logic design system, is described and an example of its use is given. SCALD provides a total computer-aided design environment which inputs a high-level description of a digital system, and produces output for computer-aided manufacture of the system. SCALD has been used in the design of an operational, 15-MIPS, 5500-chip ECL-10K processor.

STICKS - A graphical compiler for high level LSl design.

Abstract: The need for good design automation in the area of integrated circuit layout is severe. It is believed that the STICKS system does much to fill that need. STICKS is a computer aided design system which frees the designer from the tedious aspects of IC design and allows him to concentrate on the more creative and necessarily human side of the design process. With STICKS the designer is allowed to divorce himself from the usual precise sizes, spacings, and interrelationships required for an IC layout and, instead, submit his creativity in the form of rough, freehand sketches. The computer makes whatever adjustments are necessary in order to generate an error-free layout. The opportunity is then given to verify and, if necessary, modify the computer's interpretation of the design .. An interactive loop is thus formed with the computer doing all of the tedious, detailed work and the human designer providing the necessary inspiration. Using traditional' methods, large scale integrated circuit layout is a tedious, time consuming and error prone process. IC layout is a procedure which results in the placement and interconnection of each of the thousands of components which . form the integrated circuit. The designer accomplishes this by drawing several superimposed "mask" layers using a drafting board which together, control the areas where various chemical diffusions and etchings will be made on the silicon wafer that forms the IC. Components such as transistors, diodes, resistors, and their interconnections are formed by various combinations of these masked areas. Unfortunately, the specific sizes of and spacings between components is critical, preventing the designer's job from being an easy one. There is an overall goal in IC design to pack as much circuitry as possible into a small area. The more circuitry a particular IC contains, the more valuable it is and, consequently, the more marketable it is. The physical size of the chip, however, is a major factor constraining the number of components. Due to a parameter known as defect density, the larger the chip, the less likely it is to work and the more it will cost. Defect density is the number of random material defects per given area which result both from the starting silicon material and from normal losses in the manufacturing process. The solution, one would think, would be to simply make everything smaller thereby increasing component den-

A Technology-Relative Computer-Aided Design System: Abstract Representations, Transformations, and Design Tradeoffs

Abstract: This paper discusses the approach to global optimization taken in the C-MU register-transfer level computer-aided design system, currently under development at Carnegie-Mellon University, and gives examples of its application.

Register-Transfer Level Digital Design Automation: The Allocation Process

Abstract: This paper presents a portion of the register-transfer level computer aided design (RT-CAD) research at Carnegie-Mellon University. This part of the research involves the design and construction of an allocator, consisting of a set of algorithms and data structures which synthesize hardware at the logical level from a behavioral description. Preliminary results indicate the allocators performance compares favorably with a human designer.

The Maxc Systems

Abstract: The development odfthis timesharing led to insights on micrporogrammable organization, instruction sets, reliabiilty, and software and firmware development tools.

Logic Design Automation of MOS Combinational Networks with Fan-In, Fan-Out Constraints

Abstract: With the increasing complexity of current digital systems, many aspects have to be considered in the Logic design of these systems. Among them are the pin limitations, interconnections, power dissipation, testability, and fan-in, fan-out constraints. Since MOS (metal oxide semiconductor) is playing a major role in large scal integration (LSI), the problem of synthesis of MOS switching networks with large number of variables is very important. The main advantage of MOS over other forms of integrated circuits is the possibility of obtaining higher chip logic density. Long propagation delay limits MOS to low speed applications, however, speed can be improved by using advanced types of MOS as VMOS, DMOS and HMOS. An efficient algorithm for computer-aided synthesis of MOS combinational logic networks was presented in the 1977 Design Automation Conference. That algorithm considered some of the aspects need to be considered in the logic design process. These were pin limitations, interconnections, and testability. In this paper we present a new algorithm which incorporates, in addition to the aspects considered above, the fan-in, fan-out constraints. Designing digital systems with fan-in, fan-out constraints is vital for achieving proper operation of these systems - specifically speed and noise margin. A step-by-step description as well as a detailed synthesis example are given. Some of the steps of this algorithm were programmed in FORTRAN V on a SPERRY UNIVACTM 1108 computer. The effects of this kind of design on the network performance and testability are discussed. Finally, some future research problems in this area are presented.

Automatic microwave network measurements

Abstract: Network measurements using scattering parameters at microwave frequencies date back to before World-War II in different forms. With the advent of sweepers, the first phase of automation was achieved. Errors in reflection and transmission measurements were reduced by means of better hardware design until the next step of automation took place. Digital computers not only increased the speed of measurements, but were capable of correcting vectorial errors. The chronological development of S parameter-measurement techniques and error analysis is discussed.

A Minicomputer Based Interactive Graphics System as Used for Electronic Design and Automation

Abstract: Described in this paper is the use of a commercially available Interactive Graphics System for electronic design, drafting and documentation. The system is used starting from a rough schematic and ending with a tested, manufactured printed circuit board.Discussed in the paper are major aspects and quantitative results obtained in a range of computer aided tasks. These range from initial P.C. design and prototyping to final production, documentation and testing as performed in a relatively small engineering department. The turnkey system employed is Computervision's Designer TM System, with output to Computer Automation's Capable TM printed circuit board tester.Major uses of the system include its use in a novel way at the prototyping stage to produce P.C. prototype circuit boards. Others include the use of its integral minicomputer based P.C. board routing programs supplemented by interactive graphics as well as the use of a series of computer based design verification features both for checking physical constraints and for ensuring data integrity and consistency within the entire documentation package. These verifications eliminate many forms of manual checking. Actual examples are presented.

Topics in Design Automation Data Bases

Abstract: Within the last few years more and more organizations are using data base technology to manage large amounts of information. Design automation organizations are no exception. This paper discusses the application of data base technology to the field of design automation. This includes an examination of file versus data base management systems; control and integrity of design automation data; and the impact of data bases on design automation and corporate organization.

Tools for the Automation of Database Design

Abstract: In the last four years we have implemented tools for various tasks associated with database design: logical design, physical design, and re-design. An integrated example is presented to review and illustrate the capabilities of these tools. Plans for integrating the tools themselves to provide a more complete and unified service to the database designer are also presented.

The design of a dense PCB using an interactive DA system

Abstract: Historically, there have been two approaches to the use of computers for the design of printed circuit boards: Computer Aided Design (CAD) wherein a designer carries out the design of the board and a computer, coupled with graphics hardware (usually a storage tube) and software is used to carry out graphic handling and bookkeeping tasks; and Design Automation (DA) wherein a sophisticated placement and routing program, usually installed on a large computer, is used in a batch process mode to carry out the design according to necessarily limited algorithms When either of these approaches is used in conjunction with dense PCB designs, the results have been unsatisfactory: From CAD because of a feeling that "the computer should be able to help me more in solving this knotty problem"; and from DA because the results of the automated process are usually so poor as to approach being worthless: that is, the effort to complete the design approaches the level of difficulty of designing the card manually from the beginning The use of a computer-based design system for the design of dense printed circuit boards (PCB's) has thus been an elusive goal

The Design of a Data Base Organization for an Electronic Equipment DA System

Abstract: This paper presents the design of a data base organization for an electronic equipment design automation system. First, the requirement for the data base file will be summarized. Then, the logical structure the authors see to be the solution to this particular problem will be developed and demonstrated. Topics on the physical implementation of the file are presented next. Performance analyses are illustrated, and the limitations of the present state of the art in data base technique will be summarized in the conclusion.

Interactive Specification of Structured Designs

Abstract: In structured design a design is expressed in high-level components, which are defined in terms of lower-level components. Structured design combines design information of all levels in one consistent data structure. It promotes a top down design and a bottom-up implementation. High-level components are partitioned into more and more detailed components, whereas low-level components are used to build more and more complex building blocks.The components of a design communicate only through well defined interfaces and are otherwise isolated from each other. This increases the modularity of a design, and is a first step towards technology-independent design. A design can be recompiled as new design primitives become available. Capturing the structure of the design also facilitates machine processing, such as logic simulation and verification, automatic placement and layout.Structured design greatly improves the ability to understand and maintain a design. It increases the correctness of a design, it makes the designer more productive and shortens the design cycle. Finally, the design itself serves as accurate documentation.

The Effect of LSI Technology on the Theory of Modular Computer Design

Abstract: Logic-oriented and computation-oriented approaches to design offer insights into the problems of modularity.

Classification of PCB Types for Cost Effective Solutions

Abstract: Printed Circuit Board implementation has been described as the most expensive and time-consuming effort in the product design cycle Experience has proven modern design automation systems are capable of significant reductions in design time This does not say that all PCB design should be done on design automation systems This paper addresses three classes of PCB's, and describes proven techniques for methods of classification and cost effective solutions The oral presentation will be accompanied by visual aids (slides) of actual designs in each classification

Flexible models-a compromise between simplicity and accuracy in active devices modeling [MOST model for circuit CAD]

Abstract: Describes a new way to reach a compromise between simplicity and accuracy in active device modeling for computer-aided circuit design. In contrast to the conventional approach founded on the use of preliminary optimized `fixed-structure' models, a method is introduced based on consistent model optimization during the circuit analysis procedure. For the application of the method it is necessary to develop a family of recursive models with different accuracy for each kind of active device and to establish an algorithm of model simplification in accordance with a prescribed accuracy. A family of models is presented as well as one of the possible optimization procedures for MOS transistor circuits. The same method may be used for the analysis of circuits with other active devices. The method helps to gain significant CPU time economy without loss of accuracy in comparison with the conventional approach. This procedure may be easily inserted into any of the existing computer-aided design programs.

Conference Report: Hierarchical Design Stressed at Design Automation Conference

Abstract: Until recently the general attitude of engineers to design automation was confused–they questioned its value yet deplored its limited availability. To increase the amount of DA knowledge in the public domain, the 15th in the series of Annual Design Automation Conferences was held June 19-21 at Caesar's Palace, Las Vegas. Due either to this enlightened choice of venue or to spreading paranoia over handling VLSI designs, this year's conference was extremely well attended–650 delegates, an increase of 50 percent over last year. However, for those expecting ing to learn of new tools for survival in technologies governed by Moore's Laws, this conference may have been somewhat disappointing. It was largely more of the same–PWB layout, testing, IC layout, design languages, logic design, simulation, CAM, and graphics.

Automating the Software Design Process by Means of Software Design and Documentation Language

Abstract: The Software Design and Documentation Language (SDDL) approach to software design automation consists of (1) a design and documentation language with forms and syntax that are simple, unrestrictive, and communicative, (2) a processor which can take design specifications written in the SDDL language and produce an intelligible, informative, machine reproducible document, and (3) methodology for effective use of language and processor for top-down design development. The language and processor capabilities are described and suggestions and guidelines for using SDDL are discussed.

15th design automation conference report-Dac'78

Abstract: After a steady growth over the past fourteen years, this year's 15th Design Automation Conference showed a 50 percent jump in attendance to 6q7. From this conference and similar sessions at other conferences, one might conclude that the design automation programs and systems being presented are no longer just Interesting curiosities, but rather firm requirements In the design process-not just at the large electronic firms but at most firms, large and small. was Indeed a very successful conference. Caesars Palace treated the conference and Its attendees as kings. The session rooms were quiet, the overflow hotel reservations were handled quickly and courteously. A varied, In-depth technical program was provided. On yes, enter-talnment was also available for those who had time and strength after a program which started at 7:00 e.m. with tutorials and ended with 7:00 p.m. working or common-interest sessions. But then one must remember that In Las Vegas the entertainment never stops. Steve Szygende, Conference Chairman, opened the conference by reviewing the progress of Design Automation. He Indicated that originally automated draftIng was stressed whereas today very complex software systems are Involved. Steve emphasized that there was a monumental amount of work to be done In DA. He Indicated that to run out of work to be done was not the problem, but Instead 'the problem was to determine what work should and would be done with resources available. He also noted that from his observations the number of people Involved in Design Automation had rapidly Increased, as this year~s attendence also seemed to Indicate. Even before the conference officially opened, L. Bernsteln of Bell Laboratories In an early morning tutorial, "Software Manufacturing", Indicated the success that hardware design automation had achieved. Bernstein stated that software production should take cues from hardware design and manufacturing automation. These automation areas should Include source code control systems, automated build systems, and "assembly line" testing. In the tutorial, a system development cycle was detailed along with software manufacturing tasks which might lead to systematizing the production of software. Bernsteln emphasized that techniques used In the design of hardware to Increase design and build productivity should also be adopted for software. Session 2 "Computer-Aided Manufacturing at GTE Automatic Electric" further described the changes that appeared to be taking place along with the success stories In the design automation field. According to H. Olson of GTE Automatic Electric, after a highly static period …

Specification, Simulation and Automated Design of Interfaces and Digital Circuits.

Abstract: : This report describes research done in hardware description, simulation, and design automation. Although the basic thrust of the work has been aimed at I/O and interface problems, most of the results are more general. The efforts in formal hardware description have produced a language for bus, I/O and interface specification, GLIDE. GLIDE is supported by a compiler which performs syntactic and semantic checks. A translator to the ISPL language has also been written. The resulting code and non-translatable GLIDE semantics forced abandonment of ISPL either directly or indirectly from GLIDE for I/O description. The major research reported here is design automation work. This grant has supported synthesis research - the mapping from a functional description of the system (ISP) to be designed to the structure. Automated design of the control circuitry is in the early stages, but a working program designs the data paths, registers and memories. An early design the program produced came within 25% of the cost of the design a human designer produced. The PDP-8/E design, included here, has a chip count within 50% of that commercial design, but the design program produced an entirely different design.

A digital design automation course for logic designers

Abstract: This paper describes a course which is being offered to graduate students at Stanford University as well as to practicing logic designers from industry. The main purpose of this course is to familiarize logic designers with the current state-of-the-art of digital design automation.

Automated Techniques for Product-Grading Systems Design

Abstract: The study presents innovative pattern recognition techniques for design automation of systems capable of on-line supervision of product-grading activities in production environments.

Design Automation Of Centrifugal Compressors

Abstract: Tadashi Kaneki is an engineer in the centrifugal compressor design section of Tsuchiura Works, Hitachi, Ltd., Japan. He has been associated with design and development of centrifugal compressors for nine years. He played a most active part in the development of this design automation system. Mr. Kaneki graduated from Tokyo University, Japan, in 1967 and received an M.S. degree in Mechanical Engineering in 1969. Eiji Yokoyama is an assistant chief engineer for Tsuchiura Works of Hita­ chi, Ltd. , and is in charge of research and development of turbo compressors. He has been associated with design and development of axial and centrifugal compressors .for 20 years. Mr. Yokoyama received an M.S. degree in Mechanical Engineering .from Massachusetts Insti­ tute of Technology in 1961, and he is a member of ASME. Centrin1gal compressors used in various plants come in diverse specifications with regard to capacity, pressure and gas handled. In designing such compressors, important consid­ erations are to shorten design time and save labor as well as to meet user's requirements. In view of these needs, the authors developed a design automation system and has successfully applied the system to the design of all centrin1gal compressors for various chemical plants. Design automation of such centrif­ ugal compressors are described.