M. J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, C-21, 1972, pages 948--960.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....parallel machines we need a hierarchy of several masters. 4.2 Parallel programming paradigms Today parallel programming techniques have converged to two main paradigms. MPI [41] is a standard for message passing based parallel programs in C C and Fortran. It supports Flynn s taxonomy [37] of MIMD and provides a wide variety of message passing functionalities. MPI was already widely used before its initial standard appeared in 1995. It supports point to point and global communication based on some transport layer (e.g. Sockets) With MPI 2 [40] onesided communication and ....

M. Flynn. Some computer organizations and their effectiveness. IEEE Trans. Comput., C(21):94, 1972.

....without a clear understanding of network applications and scalability requirements. Also, the lack of standardized benchmarks has greatly impeded an objective comparison of solutions from different vendors. Most of the proposals to date advocate a multiple instruction, multiple data (MIMD [8]) approach to packet processing. An analysis of some of the successful architectures today (from Intel, IBM and Motorola) reveals the popularity of chip multiprocessing along with hardware multithreading in the data plane. The generic network processor contains several programmable micro engines ....

....We propose to eliminate hard wired coprocessors and implement their functionality using software executing on the micro engines themselves. Furthermore, in the conventional approach, each micro engine has its own instruction store and operates independently of other micro engines in classic MIMD [8] fashion [2] 3] 4] However, the microengines usually perform the same processing function on packets streaming through the switch. In other words, they execute the same instruction sequence. This behavior matches more closely with the single instruction, multiple data (SIMD [8] paradigm. At ....

[Article contains additional citation context not shown here]

Michael J. Flynn, "Some computer organizations and their effectiveness," IEEE Transactions on Computers,vol.C-21,no.9, 1972, pp. 948-960.

....algorithms for document clustering. 4.1 Background Parallel architectures are often described based on the number of instruction and data streams, namely single and multiple data and instruction streams. A complete taxonomy of different combinations of instruction streams and data was given in [12]. To evaluate the performance delivered by these architectures on a given computation the speedup measure is typically used. Speedup is defined as Ts Tp ,whereT s is the time taken by the best sequential algorithm, and T p is the time taken by the parallel algorithm under consideration. The ....

M.J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, 21(9):948--960, 1972.

.... of the construction and use of digital computers, several researchers have proposed various forms for parallel configuration employing different ways of parallelism [5] Without going into the hardware or software details, a classification scheme, now widely accepted, was proposed by Flynn [6, 7]. According to this classification the computer domain may be divided into four broad categories which are distinguished from each other by the number of instruction Streams and Data Streams that a computer handles at a particular time instant. These categories are: SISD: Single Instruction ....

M. J. Flynn, Some computer organizations and their effectiveness. IEEE Trans. Comput. 21, 948-960 (1972).

....difficulties and the problems that exist for the realization of these machines. 42,63,64,65,75,94,95,105] 7 2.3 Hardware 2.3.1 Classification of Parallel Processors The main and most general classification scheme for computers accepted, from user s point of view, is that proposed by Flytin. [ 1,2] This scheme distinguishes the Instructions of which any program consists and the Data, which the same program uses. A stream is a sequence of items (instructions or Data) as used by a computer. According to the number of Instruction streams and Data streams in a machine, Flynn proposes the ....

FLYNN M. J., "Some Computer Organizations and their Effectiveness", IEEE Transactions on Computers, Vol 21,9, 1972, pp. 948 - 960

.... categorized based on their ability to execute instructions and manipulate data streams sequentially or in parallel as: 1) Single Instruction Single Data (SISD) 2) Single Instruction Multiple Data (SIMD) 3) Multiple Instruction Single Data (MISD) and 4) Multiple Instruction Multiple Data (MIMD) [Flynn72]. Modern parallel computers follow the MIMD model. MIMD computers that have hundreds or thousands of processors are called Massively Parallel Processors (MPPs) MPPs typically have nonstandard dedicated interconnection network (ICN) which allows the processors to collaborate and facilitate ....

M. J. Flynn. "Some computer organizations and their effectiveness," IEEE Trans. Computers, 21(9):948-960, September 1972.

....for the entire array of processing units. Driven by these signals, each PE would execute the instruction on its own local data. Thus, a single stream of instructions was read by the CU and controlled multiple streams of data from the PMs. This processing model was later termed SIMD by Flynn [7]. The SIMD model produces speed up by taking advantage of data parallelism in the executing code. Data parallelism refers to the ability of multiple data items to be processed simultaneously without changing the results of the code execution from those which would be generated by processing the ....

Michael J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, C-21(8):948--960, 1972.

....fat tree. 128 viii Chapter 1 Introduction 1. 1 Massively Parallel Architectures Flynn gave an early classification of computer architectures based on whether the parallelism appears in the instruction stream, data stream or both [58]. In a SIMD (single instruction stream, multiple data streams) machine, each instruction is broadcast to all the processors; these processors execute the same operation on different data elements. Recent commercial machines that use this model include the MasPar MP 1 and MP 2 [108] the Connection ....

M. J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, 21(9):948--960, September 1972.

....Notable examples are the IBM 360 91 [10] and the RS 6000 [76] The studies on this type of machine generally show that it is possible to achieve modest speedups with a small instruction window. Flynn reports to find most of the available parallelism with a window size of 2 or 4 instructions in [56]: Under the constraint that instructions are not dispatched until all preceding conditional branches are resolved, stack 1 sizes as small 19 as 2 or 4 achieve most of the parallelism that a hypothetically infinite stack would. If control dependencies are strictly enforced, so that ....

M. J. Flynn, Some Computer Organizations and Their Effectiveness, IEEE Transactions on Computers C-21(9), 1972, pp. 948-960.

....algorithms for document clustering. 4.1 Background Parallel architectures are often described based on the number of instruction and data streams, namely single and multiple data and instruction streams. A complete taxonomy of different combinations of instruction streams and data was given in [9]. To evaluate the performance delivered by these architectures on a given computation, speedup is defined as Ts Tp , where T s is the time taken by the best sequential algorithm, and T p is the time taken by the parallel algorithm under consideration. The higher the speedup, the better the ....

M.J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, 21(9):948--960, 1972.

....hot spots in interconnection networks, including different schemes that have been proposed to alleviate the hot spot problem. We are interested in possible architectural differences in the broad class of MIMD (multiple instruction, multiple data stream) architectures, according to Flynn s taxonomy[46]. Although some of the results about network structure originally arose in the context of SIMD architectures or telephone switching, that literature will be reviewed only in the context of implementing highly parallel MIMD architectures. Network architectures may differ in topology, routing ....

Michael J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, C-21:948--960, 1972.

.... these being the Thinking Machines CM 2 series [1] the Intel iPSC 860 [27] hypercube, ORNL PVM [18] running on a Network of Workstations (NoW) and the Transtech Paramid [45] The CM 2 series is programmed using C Version 6 [17] a data parallel extension to C designed to support the SIMD [14] execution model used by the CM 2. C provides parallel data structures, together with operators to manipulate these structures. A single source file is required for the CM 2 series. The Intel iPSC 860 hypercube is programming using iPSC 860 C [28] a message passing extension to C. Programs run ....

M.J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, 21(9):948--960, September 1972.

....( 15] 16] These attempts using the inverted file model( 27] showed considerable performance, even under the above limitations and the limited power parallel environment they used. Concerning the type of parallelism, we can distinguish between the SIMD and MIMD environments (due to Flyn, [18]) Most parallel text retrieval systems ( 10] 15] 16] have been implemented on the earlier versions of the Connection Machine (CM 1 and CM 2) which are classical types of SIMD architectures using pipelined vector processors. However the last few years have seen increasing interest for ....

M. Flynn, Some Computer Organizations and their Effectiveness, IEEE Transactions on Computers, C--21, pp. 948--960.

No context found.

M. J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, C-21, 1972, pages 948--960.

No context found.

M. Flynn, "Some computer organizations and their effectiveness." IEEE Transactions on Computers, pp. 948960, Sep., 1972

No context found.

M. Flynn, "Some computer organizations and their effectiveness." IEEE Transactions on Computers, pp. 948960, Sep., 1972

No context found.

M. J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computers, 21(9):948-- 960, September 1972.

No context found.

M. J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, C-21(9):948--960, September 1972.

No context found.

M. J. Flynn, "Some Computer Organizations and Their Effectiveness," IEEE Transactions on Computers, Vol.21, No.9, September 1972, pp.948--960.

No context found.

M. J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computing, C-21(948-960), September 1972.

No context found.

M. J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, C-21(9):948-960, September 1972.

No context found.

M. J. Flynn. Some Computer Organizations and Their Effectiveness. IEEE Transactions on Computers, C-21(9):948-960, September 1972.

No context found.

M.J. Flynn. Some computer organizations and their effectiveness. IEEE Transactions on Computing, (C-21):948--960, September 1972.

No context found.

M.J. Flynn, "Some computer organizations and their effectiveness," IEEE Transactions on Computers, Vol.21, September 1972.

No context found.

Flynn, M.F. (1972). Some computer organizations and their effectiveness. IEEE Trans. on Comp., C/21, 9, 948--960.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC