13 Hits in 1.9 sec

On the application of massively parallel SIMD tree machines to certain intermediate-level vision tasks

Hussein A.H. Ibrahim, John R. Kender, David Elliot Shaw
1986 Computer Vision Graphics and Image Processing  
\Ve follow with a more detailed description of the P;ON-VO~ architecture (a prototype of which has been operational since January, 1985), and of the high-level parallel language in whIch our algorithms  ...  have been WrItten and simulated The heart of thE' paper consists of the deSCrIption and analysls of algorIthms for a representative Hough transform, and of an algorIthm for the interpretation of mOVlf  ...  Inter-PE communication in NON-VON IS supported by the I/O switch, a matnx of pass transistors that routes uata between the two internal buses of the SPE and its I/O ports.  ... 
doi:10.1016/0734-189x(86)90038-1 fatcat:rp2l6hy2lzey7cyjlzsua3zlmq

Large-scale parallel processing systems

Howard Jay Siegel, Thomas Schwederski, David G Meyer, William Tsun-yuk Hsu
1987 Microprocessors and microsystems  
The common factor among these applications is, in terms of computational requirements, a 'need for speed.'  ...  This paper is an introduction to some of the concepts involved in the design and use of large-scale parallel systems.  ...  The only way for a PE in GE to communicate with a PE in GO is by using the cube0 connection.  ... 
doi:10.1016/0141-9331(87)90325-5 fatcat:wb54jmakv5hb7k7wfkssjz3zwq

Computational models and resource allocation for supercomputers

J. Mauney, D.P. Agrawal, Y.K. Choe, E.A. Harcourt, S. Kim, W.J. Staats
1989 Proceedings of the IEEE  
The computational needs of a program must be cast in terms of the computational model supported by the supercomputer, and this must be done in a way that makes effective use of the machine's resources.  ...  There are several different architectures used in supercomputers, with differing computational models. These different models present a variety of resource allocation problems that must be solved.  ...  Because memory access on a processor is much faster than inter-PE communication, finding a data-mapping scheme that enables concurrent memory reference is crucial for fast computation.  ... 
doi:10.1109/5.48828 fatcat:vhlqy2v3trfwblcqpsol2zivpi


Howard Jay Siegel, Thomas Schwederski, Nathaniel J. Davis, James T. Kuehn
1984 SIGARCH Computer Architecture News  
PASM is a multifunction p_o,r titionable ~IMDfkIIMD system being designed at Purdue for parallel image understanding.  ...  PASM is to serve as a vehicle for studying the use of parallelism for performing the numeric and symbolic processing needed for tasks such as computer vision.  ...  for inter-PE data communication rather than both inter-PE communication and instruction fetch operations.  ... 
doi:10.1145/859558.859560 fatcat:r255ff2sp5aspfpgdrun2kucki

P4 [chapter]

2013 Encyclopedia of Operations Research and Management Science  
Such an arrangement helps in maintaining a sparse inverse, given that the original data set for the associated linear-programming problem is sparse.  ...  A procedure for arranging the basis matrix of a linear-programming problem into as near a lower triangular form as possible.  ...  In SIMD architectures, a single instruction stream controls all the ALUs in a synchronous manner.  ... 
doi:10.1007/978-1-4419-1153-7_200569 fatcat:aqo3msrrzree3aiyv36fsh3afe

SIMD and MSIMD Variants of the NON-VON Supercomputer

David Elliot Shaw, Columbia University. Computer Science
The NOH-VON 1 and NOS-VOH 3 machines, which are in various stages of construction at Columbia University, function for the most part as single instruction stream, multiple data stream (SIMD) [2] machines  ...  Each member of the HOH-VOH family of supercomputers employs massive parallelism to obtain substantial performance improvements in a wide range of applications.  ...  communication) far belOW tbe estimated two aicroseconds proJected for NON-VON and 3 using a nuaber of engineering techniques tbat are beyond the scope of this paper.  ... 
doi:10.7916/d86m3fz5 fatcat:cjcipe634bh3fjr6v7poq7lxea

A multimedia co-processor architecture for real-time video coding

Don Hoi Ng
Overall, the multimedia co-processor is a tightly-coupled system consisted of a specialized processing engine, an efficient memory architecture and a flexible controller all working in a coherent manner  ...  Such processing engine features a parallel architecture with specialized functional units, supporting SIMD executions.  ...  17 MIMD architecture 18 MSIMD architecture 19 Architecture of TMS320C6201 24 Block diagram of TM1000 (TriMedia) 25 SGS-Thomson Videophone Processor System 28 IMAGE chip architecture 29  ... 
doi:10.14288/1.0065274 fatcat:5hseeqiprnhnhkfft2bnpur354

Kestrel: design of an 8-bit SIMD parallel processor

D.M. Dahle, J.D. Hirschberg, K. Karplus, H. Keller, E. Rice, D.H. Speck, D.H. Williams, R. Hughey
Proceedings Seventeenth Conference on Advanced Research in VLSI  
Because computational biology (and indeed any field) is constantly discovering new analysis methods, programmability is a key issue in designing a system that can serve the needs of this vast community  ...  world: data input and output (I/O), inter-PE communication, instruction broadcast, and cooperation with the host computer. 2.1: Data Input and Output Kestrel is primarily a linear array of processing elements  ...  Thus, a good inter-PE communication scheme is essential for any array processor. Kestrel employs Syst,olic Shared Registers (SSRs) for inter-PE communication ( Figure 1 ).  ... 
doi:10.1109/arvlsi.1997.634852 dblp:conf/arvlsi/DahleHKKRSWH97 fatcat:e3kb5nzjo5gt3m6rlhzomhml64

DADO: A Parallel Processor for Expert Systems

Salvatore Stolfo, Daniel P. Miranker, Columbia University. Computer Science
DADO is a parallel, tree-structured machine designed to provide significant performance improvements in the execution of large expert systems implemented in production system form.  ...  The PE's are introduced to form a complete binary tree. This paper describes the application domain of the DADO machine and the rationale for its design.  ...  n point in time. Thi3 nexibl. architectural dtsign supports multiple-ST.\{[) execution (MSIMD). as. for example, 12tl. but on a much larger scal..  ... 
doi:10.7916/d8f196vh fatcat:nopkr3doxbfxnkxops7flp7ztm

The DADO Parallel Computer

Salvatore Stolfo, Columbia University. Computer Science
DADO is a parallel, tree-structured machine designed to provide significant performance improvements in the execution of large production systems.  ...  The PE's are interconnected to form a complete binary tree. This paper describes the organization of, and programming language for two prototypes of the DADO system.  ...  xible architectural design supports mult£ple-SlAfD execution (MSIMD). Thus. thp m<J.  ... 
doi:10.7916/d8vm4m73 fatcat:jpa2by3bszauvm6nsnxxrefwau

Image Understanding Algorithms on Fine-Grained Tree-Structured SIMD Machines

Hussein A. H. Ibrahim
· have led to several proposals for parallel architectures for computer vision systems.  ...  Such systems typically require a large amount of computation for processing raw Image data at the lowest level, and for sophisticated decision making at the highest level Recent advances In VLSI circuitry  ...  The MIMD, SIMD, and MSIMD capablhties of the proposed architecture and the aVailability of fine-and medium-grained PE's should prove useful In implementing systems that perform well on all levels of computer  ... 
doi:10.7916/d8zc8bcm fatcat:erby2x4gkng2rdeafeyaqcvjpi


Uredniski Odbor, Aleksic, Bitraikov, Skopje, Rijeka Dragojlovle, Ljubljana Hodzar, Horvat, Sarajevo Mandzic, Mihalie, Turk
A modern CAD system has his awn program's inter-face.  ...  In particular VLSI systems offer new scope in computing for designing dedicated architectures for a variety of applications.  ...  This article presents a multl tasking environment for the MS-DOS operating system on the IBH-PC computer.  ... 

Solving the Depth Interpolation Problem on a Parallel Architecture with Efficient Numerica1 Methods

Dong Jae Choi, Columbia University. Computer Science
We solve the depth interpolation problem on a parallel architecture, a fine grained SIMD machine with local and global communication networks.  ...  Many constraint propagation problems in early vision, including depth interpolation, can be cast as solving a large system of linear equations where the resulting matrix is symmetric and positive definite  ...  data between the two interna1 buses and由e three inter-PE communication buses (parent, left child, and right child) in 由e course of executing inter-PE communication instructions.  ... 
doi:10.7916/d8j391pt fatcat:iyg6vzarrvbw3io4bndlytgc7y