Search
Join ACM

Architecture and Hardware

Advertisement

Research and Advances Mar 1 1978

Implications of structured programming for machine architecture

Based on an empirical study of more than 10,000 lines of program text written in a GOTO-less language, a machine architecture specifically designed for structured programs is proposed. Since assignment, CALL, RETURN, and IF statements together account for 93 percent of all executable statements, special care is given to ensure that these statements can be implemented efficiently. A highly compact instruction encoding scheme is presented, which can reduce program size by a factor of 3. Unlike a Huffman code, which utilizes variable length fields, this method uses only fixed length (1-byte) opcode and address fields. The most frequent instructions consist of a single 1-byte field. As a consequence, instruction decoding time is minimized, and the machine is efficient with respect to both space and time.
Research and Advances Jan 1 1978

The evolution of the DECsystem 10

The DECsystem 10, also known as the PDP-10, evolved from the PDP-6 (circa 1963) over five generations of implementations to presently include systems covering a price range of five to one. The origin and evolution of the hardware, operating system, and languages are described in terms of technological change, user requirements, and user developments. The PDP-10's contributions to computing technology include: accelerating the transition from batch oriented to time sharing computing systems; transferring hardware technology within DEC (and elsewhere) to minicomputer design and manufacturing; supporting minicomputer hardware and software development; and serving as a model for single user and timeshared interactive minicomputer/microcomputer systems.
, , , and
Research and Advances Jan 1 1978

The development of the MU5 computer system

Following a brief outline of the background of the MU5 project, the aims and ideas for MU5 are discussed. A description is then given of the instruction set, which includes a number of features conducive to the production of efficient compiled code from high-level language source programs. The design of the processor is then traced from the initial ideas for an associatively addressed “name store” to the final multistage pipeline structure involving a prediction mechanism for instruction prefetching and a function queue for array element accessing. An overall view of the complete MU5 complex is presented together with a brief indication of its performance.
and
Research and Advances Jan 1 1978

The Manchester Mark I and atlas: a historical perspective

In 30 years of computer design at Manchester University two systems stand out: the Mark I (developed over the period 1946-49) and the Atlas (1956-62). This paper places each computer in its historical context and then describes the architecture and system software in present-day terminology. Several design concepts such as address-generation and store management have evolved in the progression from Mark I to Atlas. The wider impact of Manchester innovations in these and other areas is discussed, and the contemporary performance of the Mark I and Atlas is evaluated.
Research and Advances Jan 1 1978

The CRAY-1 computer system

This paper describes the CRAY-1, discusses the evolution of its architecture, and gives an account of some of the problems that were overcome during its manufacture. The CRAY-1 is the only computer to have been built to date that satisfies ERDA's Class VI requirement (a computer capable of processing from 20 to 60 million floating point operations per second) [1]. The CRAY-1's Fortran compiler (CFT) is designed to give the scientific user immediate access to the benefits of the CRAY-1's vector processing architecture. An optimizing compiler, CFT, “vectorizes” innermost DO loops. Compatible with the ANSI 1966 Fortran Standard and with many commonly supported Fortran extensions, CFT does not require any source program modifications or the use of additional nonstandard Fortran statements to achieve vectorization. Thus the user's investment of hundreds of man months of effort to develop Fortran programs for other contemporary computers is protected.
Research and Advances Jan 1 1978

Architecture of the IBM system/370

This paper discusses the design considerations for the architectural extensions that distinguish System/370 from System/360. It comments on some experiences with the original objectives for System/360 and on the efforts to achieve them, and it describes the reasons and objectives for extending the architecture. It covers virtual storage, program control, data-manipulation instructions, timing facilities, multiprocessing, debugging and monitoring, error handling, and input/output operations. A final section tabulates some of the important parameters of the various IBM machines which implement the architecture.
and
Research and Advances Jan 1 1978

The evolution of the Sperry Univac 1100 series: a history, analysis, and projection

The 1100 series systems are Sperry Univac's large-scale mainframe computer systems. Beginning with the 1107 in 1962, the 1100 series has progressed through a succession of eight compatible computer models to the latest system, the 1100/80, introduced in 1977. The 1100 series hardware architecture Is based on a 36-bit word, ones complement structure which obtains one operand from storage and one from a high-speed register, or two operands from high-speed registers. The 1100 Operating System is designed to support a symmetrical multiprocessor configuration simultaneously providing multiprogrammed batch, timesharing, and transaction environments.
, , and
Research and Advances Oct 1 1975

Merging with parallel processors

Consider two linearly ordered sets A, B, | A | = m, | B | = n, m ≤ n, and p, p ≤ m, parallel processors working synchronously. The paper presents an algorithm for merging A and B with the p parallel processors, which requires at most 2⌈log2(2m + 1)⌉ + ⌊3m/p⌋ + [m/p][log2(n/m)] steps. If n = 2&bgr;m (&bgr; an integer), the algorithm requires at most 2[log2(m + 1)] + [m/p](2 + &bgr;) steps. In the case where m and n are of the same order of magnitude, i.e. n = km with k being a constant, the algorithm requires 2[log2(m + 1)] + [m/p](3 + k) steps. These performances compare very favorably with the previous best parallel merging algorithm, Batcher's algorithm, which requires n/p + ((m + n)/2p)log2m steps in the general case and km/p + ((k + 1)/2)(m/p)log2m in the special case where n = km.
Research and Advances Aug 1 1975

Combining decision rules in a decision table

The techniques for minimizing logic circuits are applied to the simplification of decision tables by the combining of decision rules. This method is logically equivalent to the Quine-McCluskey method for finding prime implicants. If some of the decision rules implied in the ELSE Rule occur with low frequency, then the ELSE Rule can be used to further simplify the decision table. Several objectives merit consideration in optimizing a decision table: reducing machine execution time; reducing preprocessing time; reducing required machine memory; reducing the number of decision rules. (This often improves the clarity of the decision table to a human reader.) It will be shown that objectives (3) and (4) can be furthered with the above methods. Objective (1) is also attained if overspecified decision rules are not combined. Objective (2) must be compared against the potential benefits of objectives (1), (3), and (4) in deciding whether to use the above methods.
Research and Advances Mar 1 1975

Glypnir—a programming language for Illiac IV

GLYPNIR is one of the earliest existing languages designed for programming the Illiac IV computer. The syntax of the language is based on ALGOL 60, but has been extended to allow the programmer explicitly to specify the parallelism of his algorithm in terms of 64-word vectors. This paper describes the characteristics, goals, and philosophy of the language, and discusses some of the problems associated with parallel computer architectures.
, , , and
Research and Advances Jul 1 1974

Formal requirements for virtualizable third generation architectures

Virtual machine systems have been implemented on a limited number of third generation computer systems, e.g. CP-67 on the IBM 360/67. From previous empirical studies, it is known that certain third generation computer systems, e.g. the DEC PDP-10, cannot support a virtual machine system. In this paper, model of a third-generation-like computer system is developed. Formal techniques are used to derive precise sufficient conditions to test whether such an architecture can support virtual machines.
and
Research and Advances Jul 1 1974

The UNIX time-sharing system

UNIX is a general-purpose, multi-user, interactive operating system for the Digital Equipment Corporation PDP-11/40 and 11/45 computers. It offers a number of features seldom found even in larger operating systems, including: (1) a hierarchical file system incorporating demountable volumes; (2) compatible file, device, and inter-process I/O; (3) the ability to initiate asynchronous processes; (4) system command language selectable on a per-user basis; and (5) over 100 subsystems including a dozen languages. This paper discusses the nature and implementation of the file system and of the user command interface.
and
Research and Advances Apr 1 1974

Execution characteristics of programs in a page-on-demand system

h show the execution characteristics of two types of commonly used programs in a large-scale, time-shared computer system. A software monitoring facility built into the supervisor was used for data collection during normal system operation. These data were analyzed, and results of this analysis are presented for a Fortran compiler and an interactive line file editor. Probability distribution functions and other data are given for such things as CPU intervals, I/O intervals, and the number of such intervals during execution. Empirical distributions are compared with simple theoretical distributions (exponential, hyperexponential, and geometric). Other data show paging characteristics of tasks as a function of the number of pages those tasks have in core.
Research and Advances Apr 1 1974

A simple linear model of demand paging performance

Predicting the performance of a proposed automatically managed multilevel memory system requires a model of the patterns by which programs refer to the information stored in the memory. Some recent experimental measurements on the Multics virtual memory suggest that, for rough approximations, a remarkably simple program reference model will suffice. The simple model combines the effect of the information reference pattern with the effect of the automatic management algorithm to produce a single, composite statement: the mean number of memory references between paging exceptions increases linearly with the size of the paging memory. The resulting model is easy to manipulate, and is applicable to such diverse problems as choosing an optimum size for a paging memory, arranging for reproducible memory usage charges, and estimating the amount of core memory sharing.
Research and Advances Mar 1 1974

Scan conversion algorithms for a cell organized raster display

Raster scan computer graphics with “real time” character generators have previously been limited to alphanumeric characters. A display has been described which extends the capabilities of this organization to include general graphics. Two fundamentally different scan conversion algorithms which have been developed to support this display are presented. One is most suitable to noninteractive applications and the other to interactive applications. The algorithms were implemented in Fortran on the CDC6400 computer. Results obtained from the implementations show that the noninteractive algorithms can significantly reduce display file storage requirements at little cost in execution time over that of a conventional raster display. The interactive algorithm can improve response time and reduce storage requirements.
and
Research and Advances Mar 1 1974

Optimal space allocation on disk storage devices

When the amount of space required for file storage exceeds the amount which can be kept online, decisions must be made as to which files are to be permanently resident and which mountable. These decisions will affect the number of mount requests issued to the operators. This is often a bottleneck in a computing facility, and reducing the number of mounts thus decreases turnaround time. An optimization model for the assignment of files to disk packs, and packs to either resident or nonresident status is presented. Heuristics are suggested for those cases in which it is Inefficient to compute the actual optimum.
Research and Advances Mar 1 1974

On the construction of a representative synthetic workload

A general method of constructing a drive workload representative of a real workload is described. The real workload is characterized by its demands on the various system resources. These characteristics of the real workload are obtained from the system accounting data. The characteristics of the drive workload are determined by matching the joint probability density of the real workload with that of the drive workload. The drive workload is realized by using a synthetic program in which the characteristics can be varied by varying the appropriate parameters. Calibration experiments are conducted to determine expressions relating the synthetic program parameters with the workload characteristics. The general method is applied to the case of two variables, cpu seconds and number of I/O activities; and a synthetic workload with 88 jobs is constructed to represent a month's workload consisting of about 6000 jobs.
and
Research and Advances Feb 1 1974

A cell organized raster display for line drawings

Raster scan computer graphics displays with “real time” character generators have previously been limited to alphanumeric characters. A display is described which extends the capabilities of this organization to include general graphics. The feasibility of such a display is shown by deriving the minimum number of patterns required in the read only memory of the character generator to synthesize an arbitrary line. The synthesis process does not compromise picture quality, since the resulting dot patterns are identical with those of a conventional raster display. Furthermore, the time constraints of a raster display are shown to be satisfied for a typical design for very complex line drawings.
and
Research and Advances Oct 1 1969

A modular computer sharing systems

An alternative approach to the design and organization of a general purpose interactive multiterminal computing system is presented. The system organization described is a conceptually simple arrangement of a bank of interchangeable computers, each of which is a memory/proputor pair, that are assigned to process terminal jobs as they arrive. One of the computers serves as the master or control computer and supervises the collection and distribution of messages from and to the remote terminals. In the simplest form there is a disk drive for each connected terminal. A crosspoint switching network allows any such disk drive to be connected to any computer in the bank, under control of the control computer. Thus, while each active terminal user “occupies” a dedicated disk drive, he may share the computer with many other terminal users in a simple manner. The ratio of users to computers is dependent on both the size and power of the machines used and the computation requirements of the particular mix of users. This system organization is inherently a simpler and therefore more reliable approach to time-sharing computers and has the potential of a highly available system at relatively low cost. Economic configurations are possible for a range of systems sizes that span at least one order of magnitude. Finally, problem programs developed by remote terminal users can be run on a dedicated batch system if compatible computers are used.
, , , and

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More