Simulation based performance prediction by PEPSY

نویسنده

  • Roman Blasko
چکیده

Parallel programs, generated by the supercompiler, VFCS (Vienna Fortran Compilation System), are an application area for the tool, PEPSY (PErformance Prediction System), which we have developed recently. PEPSY derives a performance model from an internal representation of the parallel program in the compiler, automatically, and makes performance analysis by discrete-event simulation. Several monitoring modes of PEPSY enable t o analyze a parallel program behavior at the detailed and global level, evaluating static and dynamic performance indices. Performance prediction is cam'ed out before the final target parallel code is generated, or before a real run is executed on the parallel computer, in order to optimize the parallelization strategy for VFCS. This paper as focused on the prediction of a new subset of performance indices for the parallel programs, evaluated b y PEPSY, and illustrating its new monitoring facilities suitable for comparison of different versions of the program.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Leistungsbewertung mit PEPSY-QNS und MOSES

In this article the two tools PEPSY-QNS and MOSES, whose areas of application complement one another, are introduced. Dynamic systems, which can be described by queueing networks (like computer or manufacturing systems) can be analyzed with PEPSY-QNS (Performance Evaluation and Prediction System for Queueing Networks). To describe the system in PEPSY-QNS, the graphical user interface XPEPSY is ...

متن کامل

A Rapid Prototyping Environment for Multi-DSP Systems based on Accurate Performance Prediction

In this paper we present Pepsy, a novel prototyping environment for multiDSP systems, with the primary goal to support the design and implementation of parallel digital signal processing (DSP) applications subject to various design constraints. Given a specification of the prototyping problem in the form of an application model, a hardware model and mapping constraints, Pepsy automatically maps...

متن کامل

Prediction of Nitrogen Injection Performance in Conventional Reservoirs Using the Correlation Developed by the Incorporation of Experimental Design Techniques and Reservoir Simulation

Enhanced oil recovery using nitrogen injection is a commonly applied method for pressure maintenance in conventional reservoirs. Numerical simulations can be practiced for the prediction of a reservoir performance in the course of injection process; however, a detailed simulation might take up enormous computer processing time. In such cases, a simple statistical model may be a good approach to...

متن کامل

Neuro-Fuzzy Based Algorithm for Online Dynamic Voltage Stability Status Prediction Using Wide-Area Phasor Measurements

In this paper, a novel neuro-fuzzy based method combined with a feature selection technique is proposed for online dynamic voltage stability status prediction of power system. This technique uses synchronized phasors measured by phasor measurement units (PMUs) in a wide-area measurement system. In order to minimize the number of neuro-fuzzy inputs, training time and complication of neuro-fuzzy ...

متن کامل

Water Flooding Performance Evaluation Using Percolation Theory

Water flooding is a well-known secondary mechanism for improving oil recovery. Conventional approach to evaluate the performance of a water flooding process (e.g. breakthrough and post breakthrough behavior) is to establish a reliable geological reservoir model, upscale it, and then perform flow simulations. To evaluate the uncertainty in the breakthrough time or post breakthrough behavior, thi...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1995