Generating Finite Element Programs for Shared Memory Multiprocessors
نویسندگان
چکیده
Symbolic computation is employed to automatically derive formulas in finite element analysis (FEA) and to generate parallel numeric code. Key FEA computations parallelized include element stiffness computations and solution of global system of equations. An element-by-element preconditioned conjugate gradient method is used to solve the global system of equations in parallel. Derived formulas are automatically mapped onto the shared-memory architecture. An experimental software system, P-FINGER, is being extented. P-FINGER features a specification language to describe numeric algorithms for which code is to be generated. The specifications also allow an automatic code dependence analysis mechanism to extract parallelism from the specified computational steps. A separate code translator GENCRAY is modified to render code into parallel f77. Generated parallel routines run under the control of existing FEA packages. Examples of generated code are also presented. 1Work reported herein has been supported in part by the National Science Foundation under Grants CCR-8714836 and EET8714628
منابع مشابه
Testing Memory Consistency of Shared-memory Multiprocessors a Dissertation Submitted to the Department of Electrical Engineering and the Committee on Graduate Studies of Stanford University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
Shared-memory multiprocessors are becoming the dominant architecture for singlechip and multi-chip microprocessor based systems. Shared memory architectures are difficult to design because they must correctly implement the complexity of cache coherence and a memory consistency model. Memory consistency is a contract between hardware and software that specifies how memory behaves with respect to...
متن کاملSystem Software Support for Reducing Memory Latency on Distributed Shared Memory Multiprocessors
This paper overviews results from our recent work on building customized system software support for Distributed Shared Memory Multiprocessors. The mechanisms and policies outlined in this paper are connected with a single conceptual thread: they all attempt to reduce the memory latency of parallel programs by optimizing critical system services, while hiding the complex architectural details o...
متن کاملModeling and Performance Evaluation of Multi-Processors Organization with Shared Memories
This paper is primarily concerned with theoretical evaluation of the performance of multiprocessors system. A markovian waiting line model has been developed for various different multi-processors configurations, with shared memory. The system is analysed at the request level rather than job level.
متن کاملExecution-Driven Simulation of Shared-Memory Multiprocessors
This paper describes an eecient execution-driven technique for the simulation of shared-memory multiprocessors driven by real programs. Our simulator ooers substantial advantages in terms of reduced time and space overheads when compared to instruction-driven or trace-driven simulation techniques, without signiicant loss of accuracy. The technique produces correctly interleaved address traces a...
متن کاملPerformance Characterization of Shared- and Distributed-Memory Multiprocessors on a Tree Search Problem
In this paper, we measure and compare the performance of sharedand distributed-memory multiprocessors using a parallel tree search problem to characterize these types of multiprocessors. We take the knapsack problem using the branch-and-bound algorithm as our workload. It is di cult to compare the performance using irregular parallel problems such as tree search problems because the parallelism...
متن کامل