Motivation and Design Methodology for the Development of a Massively Parallel SIMD Architecture
نویسنده
چکیده
An SIMD massively parallel computer has to compete with workstations and parallel systems based on off-the-shelf microprocessor. This report motivate the choice for the development of a SIMD massively parallel architecture for image processing focusing on the constraints under which the architecture can succesfully compete with available sequential systems. The correct design methodology is also pointed out.
منابع مشابه
Target System Analysis Transformation and Adaption steps Hardware Specification
Based on a stepwise software development methodology in this paper a possibility is presented to formally describe the algorithmic details and hardware-characteristics innuencing eecient problem solutions for massively parallel systems. Variability at the speciication level is introduced by the possibility to express properties and modiication rules of the mathematical objects and expressions t...
متن کاملVolume Visualisation on a Hybrid Parallel Architecture
In this paper we present the design of a new hybrid system based on commodity components to gain supercomputer power at low cost. The architecture is built around a MIMD PC-cluster linked by a high-speed network and massively parallel SIMD processor boards connected to each node. We present the design of a volume rendering algorithm in order to derive an efficient mapping onto the new architect...
متن کامل3 the Basis of the Machine : the Programming Model
The design of a SIMD machine is usually complex because it leads to developping an eecient Processing Element and to writing all the softwares required by the chip and the control of the machine. We propose a diierent approach by using an eecient 32-bit oo-the-shelf processor with its software environment (compiler and assembler) and a programmable gate array for the network. It limits the deve...
متن کاملModel-Based Application Development for Massively Parallel Embedded Systems
The development of embedded systems in information-rich contexts is governed by some intertwined trends. The increase of both volume of data to be processed and the related processing functionality feeds the growing complexity of applications. Independently, the processing hardware that is needed to process these applications, is becoming increasingly parallel and heterogeneous (many-core) beca...
متن کاملThe Rewrite Rule Machine Node Architecture and Its Performance
The Rewrite Rule Machine (RRM) is a massively parallel MIMD/SIMD computer designed with the explicit purpose of supporting veryhigh-level parallel programming with rewrite rules. The RRM’s node architecture consists of a SIMD processor, a SIMD controller, local memory, and network and I/O interfaces. A 64-node cluster board is already an attractive RRM system capable of extremely high performan...
متن کامل