Dataflow Computing System Research Articles

Features of Creating Parallel Programs for the Parallel Dataflow Computing System "Buran"

The use of modern multi-core and multi-processor computer systems for actual tasks is not effective enough, even taking into account the use of parallel programming technologies. The solution to the problem of efficient loading of computing resources can be the transition to computing models and architectures that are inherently parallel. One of such architectures is the Parallel Dataflow Computing System (PDCS) "Buran", which implements a dataflow computing model with a dynamically formed context. A feature of the dataflow computing model is the activation of computations by data readiness, which affects both the architecture of the computing system and the creation of programs for such systems. Differences between imperative and dataflow programming paradigms are also reflected in the route of creating a program, especially its parallel implementation. The route of creating a dataflow parallel program differs markedly from the traditional one. Already at the first stage, a parallel algorithm is created and implemented (including the algorithm for generating initial data). Next, this algorithm is debugged when it is executed on an emulator or model of the system with one computing core. After that, the selection and configuration of function of computation distribution is performed, the program is executed (without changing its code) on the emulator or model of the system with several computing cores, and, finally, the program is debugged in multi-core mode. These stages of the route differ from similar stages of the traditional route, both in form and in essence. Unlike traditional parallel, and even more so sequential programs, it can be said that a dataflow program in equal parts consists of a program code that implements a task, an algorithm for generating initial data, and a function of computation distribution. This thesis is demonstrated by the example of solving the problem of finding sum of array elements, where the difference between the implementations is in the algorithm for generating initial data, which radically affects the nature of the dataflow program passing. Careful attention to each of the parts of the dataflow program is the key to the correct and efficient solution of problems on the PDCS, which provides support to the programmer at the hardware level.

Dynamic Control of Computation Consistency in the Parallel Dataflow Computing System

When developing high-performance multiprocessor computing systems, much attention is paid to ensuring uninterrupted operation, both in terms of hardware and software. In traditional computing systems, software is the main focus in address­ing these issues. The article discusses the solution to the issue of ensuring uninterrupted operation for the parallel dataflow computing system (PDCS), which implements the dataflow computational model with a dynamically formed context. Due to the features of the PDCS, it is proposed to implement this type of control in hardware, which will increase its efficiency, since the computational process will be controlled in dynamics, and not only in statics.

Средства верификации распределения вычислений в потоковой архитектуре ППВС «Буран»

Средства верификации распределения вычислений в потоковой архитектуре ППВС «Буран»

Organization of Computations in Conditions of Limited Amount of Content Addressable Memory in the Parallel Dataflow Computing System "Buran"

Organization of Computations in Conditions of Limited Amount of Content Addressable Memory in the Parallel Dataflow Computing System "Buran"

Genetic scheduling policy on codelet model

SummaryThe Codelet Model is a fine‐grained event‐driven hybrid parallel model inspired by dataflow, whose computing performance depends on the scheduling policy. An approximate optimal codelet scheduling policy based on the features of the task graphs is important to accelerate the performance of dataflow computer system. Therefore, we have proposed an adaptive genetic scheduling policy (GSP) for codelet by improving a “pure” genetic algorithm (IPGA) for given tasks with complex dependencies. It is verified that the genetic scheduling policy is effective according to bunches of experimental results.

Методы планирования вычислений в параллельной потоковой вычислительной системе «Буран»

Методы планирования вычислений в параллельной потоковой вычислительной системе «Буран»

Разработка и исследование алгоритма задачи перемножения разреженных матриц для параллельной потоковой вычислительной системы «Буран»

Разработка и исследование алгоритма задачи перемножения разреженных матриц для параллельной потоковой вычислительной системы «Буран»

Topological optimization of the large-scale data transmission networks

Consideration was given to the design of the topological structure of the doubly connected large-scale networks by the cost criterion and constraints on the diameter. An exact combinatorial algorithm to solve the problem and a parallel algorithm to carry out calculations on the multiprocessor data-flow computer system were proposed. Parallelization was shown to provide a substantial gain in the time of computations in the optimal topological structure of the large-scale networks.

Two Tools for Interprocess Communication in Distributed Data-Flow Systems

Existing data-flow languages are incapable of supporting large, manageable and reliable non-deterministic programs. Operating system implementations and real-time applications require high-level non-deterministic programming facilities. Hoare's Communicating Sequential Processes and Brinch Hansen's Distributed Processes are two widely acclaimed distributed programming languages for conventional distributed computing systems. The powerful constructs of interprocess communication and synchronisation supported by the above two languages have been incorporated in Communicating Data Flow Channels (CDFC), and Distributed Data Flow Channels (DDFC), two languages proposed in this paper for distributed processing in data-flow systems. Several new features have been incorporated in these languages to make their execution more efficient in a data-flow environment. The implementation of these two languages on an abstract data-flow machine is investigated in detail. Two simulators have been implemented for testing the effectiveness of our implementation schemes for these two data-flow languages. These simulators are written in Pascal and run on a DEC 1090 computing system. This work demonstrates that a data-flow computing system can accommodate, without taking resort to specialised hardware units, the non-deterministic and synchronisation constructs of conventional languages with the attendant benefits of parallel data-driven execution.

The Manchester dataflow machine

A prototype dataflow computer system has been constructed by a search team at the University of Manchester. The hardware has been operational since October 1981, but has been steadily enhanced since that time. Store capacities and I/O bandwidth are approaching the state where realistically large application programs can be used to evaluate system performance. During the period of hardware enhancement, there have been parallel advances in the development of dataflow system software in the form of assemblers, compilers, debugging systems and sundry software tools. The software is also approaching readiness for application to large-scale benchmark programs. The Manchester system implements a tagged-token dataflow model of computation. This model imposes a tag-field penalty on data values in order to maximise asynchronousness of instruction execution. It is important to know to what extent this overhead is necessary, and how to minimise it whilst maintaining acceptable asynchronousness. In comparison with more conventional architectures it is important that useful measures of cost and performance be developed. With emphasis on these issues, the paper describes the structure of the Manchester dataflow hardware and software, and outlines the system performance results so far obtained. Whilst this work is far from complete, it suggests new avenues for hardware and software development which are being followed at Manchester and elsewhere.

The Manchester dataflow machine

A prototype dataflow computer system has been constructed by a research team at the University of Manchester. The hardware has been operational since October 1981, but has been steadily enhanced since that time. Store capacities and I/O bandwidth are approaching the state where realistically large applications programs can be used to evaluate system performance. During the period of hardware enhancement, there have been parallel advances in the development of dataflow system software in the form of assemblers, compilers, debugging systems and sundry software tools. The software is also approaching readiness for application to large-scale benchmark programs. The Manchester system implements a tagged-token dataflow model of computation. This model imposes a tag-field penalty on data values in order to maximise asynchronousness of instruction execution. It is important to know to what extent this overhead is necessary, and how to minimise it whilst maintaining acceptable asynchronousness. In comparison with more conventional architectures it is important that useful measures of cost and performance be developed. With emphasis on these issues, the paper describes the structure of the Manchester dataflow hardware and software, and outlines the system performance results so far obtained. Whilst this work is far from complete, it suggests new avenues for hardware and software development which are being followed at Manchester and elsewhere.

Implementing streams on a data flow computer system with paged memory

In several data flow architectures, “ streams ” are proposed as special data structures able to improve parallel execution in functional programs by providing a pipelining effect between different program parts. This paper describes how streams are implemented on a data flow computer system based on a paged memory. This memory holds both the data flow programs and data structures such as streams. Streams are stored in the memory as a linked list of pages while pointers to the streams are flowing as data tokens. A reference count is used to prevent for excessive copying of data and to control the allocation and recovery of pages. Input/output is treated as a special application of streams.

Real Time Multi-Microprocessor System Introduced Concept of Data-Driven

A multi-microprocessor system -GMMCS-with a time slice common bus is built at university of Gunma. Concept of data driven is introduced into its control or operating program. This introduction makes the operating program simple and clear, and the system can work as a procedure level data flow computer system with plural master processors. This paper discrives a method of introducing the concept into the system and some executing results of programs.

Object recognition using a data-flow computing system

Considerations of cost, performance, reliability and flexibility are leading to the increasing decentralisation of computing systems in control applications. A decentralised computing system based on the data-flow model of computation is applied, in conjunction with a laser range-finder, to the task of recognising a number of simple objects in a work-space. The recognition process involves a combination of active interrogation and multiple-hypothesis testing. Processing-elements in the system need not be identical and may be based on conventional microprocessors.

Real time multi-microprocessor system introduced concept of data-driven

A multi-microprocessor system -GMMCS- with a time slice common bus is built at university of Gunma. Concept of data driven is introduced into its control or operating program. This introduction makes the operating program simple and clear, and the system can work as a procedure level data flow computer system with plural master processors. This paper discrives a method of introducing the concept into the system and some executing results of programs.