Multiprocessor implementation of algorithms for ordinary differential equations

Article Abstract:

Multiprocessor computers constitute a cost-effective alternative to supercomputers for applications that require the solution of large sets of ordinary differential equations. The efficient use of multiprocessor computers however, requires tools for generation of parallel applications. This paper presents a general methodology for parallelization of the integration algorithms through a heuristic procedure for static multiprocessor scheduling of the computations involved. The procedure partitions, distributes and schedules the entire computation, based on decomposition of the mathematical models. Both problems are addressed: (i) minimize the completion time for a specified number of processors, and (ii) minimize the number of processors required to achieve a specified performance. The procedure combines high performance and simplicity for a wide variety of mathematical models, numerical algorithms and multiprocessor architectures.

Optimization theory, Optimization, Mathematics of Computing, Scheduling, Mathematical Programming, Heuristic Methods, Methods, Ordinary Differential Equations, Load Balancing, Load balancing (Computers), Parallel Algorithms

---

Dynamic nuclear power plant simulations using single-board peripheral array processors

Article Abstract:

The application of a host/coprocessor system, consisting of an IBM PC/AT and a Marinco peripheral array processor, to the simulation of nuclear power plant transients is described. The simulation algorithm employs a multistep, multirate integration method. Rapidly varying transients were implemented on the array processor, while the slow components of the system dynamics were handled by the host computer. The generation of nonlinear functions of dependent variables using table lookup techniques was also implemented on the array processor. A simplified model of a pressurized water reactor was employed as a benchmark. Flow rate and power transients using a variety of integration formulas are presented along with data showing the speedup achieved through the use of the peripheral processor. (Reprinted by permission of the publisher.)

Nuclear power plants, Coprocessors, Modeling, Data modeling software, Microcomputer, Scientific Research, Array Processor, Add-In/On Devices, Single-Board Computer, Coprocessor

---

ALI: a CSSL-multiprocessor software interface

Article Abstract:

ALI (A Language Interface) is a software package which translates simulation models expressed in one of the higher-level languages, CSSL-IV or ACSL, into sequences of instructions for each processor of a network of microprocessors. The partitioning of the source program among the processors is automatically accomplished. The code is converted into a data flow graph, analyzed and divided among the processors to minimize the overall execution time in the presence of interprocessor communication delays. The paper describes ALI from the user's point of view and includes a detailed example of the application of ALI to specific dynamic system simulation. (Reprinted with permission of the publisher.)

Software, Microprocessor, CPUs (Central processing units), Interface, Models, Code Conversion, Data Flow Diagrams

Similar abstracts:

• Abstracts: The choice of replenishment policies in an MRP environment. A framework for cost-service analysis in differentiation of logistics services
• Abstracts: Improve distillation column control. A new option: osmotic distillation
• Abstracts: Career orientations and career success among research, and development and engineering professionals. An investigation of factors affecting how engineers and scientists seek information
• Abstracts: Creating a sensor of achievement. Drives gain riches from the deep. (motors and drives). All quiet for reluctant motors
• Abstracts: Optimal just-in-time buffer inventory for regular preventive maintenance. Optimal control of preventive maintenance schedule and safety stocks in an unreliable manufacturing environment

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.