J. C. Yan, S. R. Sarukkai, and P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit," Software Practice & Experience, Vol. 25, No. 4, pp. 429-61.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....structure level, as specified by the programmer. 1. Introduction Understanding and tuning memory system performance is a critical issue for most scientific programs. To help programmers tune their programs, a variety of tools have been created ranging from source code and binary analysis tools [ 1, 2, 3, 4, 5] to libraries and utilities to access hardware performance counters built into microprocessors [6, 7, 8, 9] Depending on the type of problem being studied and stage of the tuning process (initial tuning of a new algorithm vs. frae tuning for a specific platform) different tools are useful. One ....

J. Yan, S. Sarukkai, and P. Mehra. "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs Using the AIMS Toolkit". Software Practice & Experience, 25(4) pages 429-461, April 1995.

....toolkit These tools integrate monitoring and statistical modeling techniques. Measurements are used to parameterize the model, which is subsequently used for predicting performance. The IS performs the basic data collection tasks. Reference: http: www.nas.nasa.gov NAS Tools Projects AIMS and [4,33] Performance and Program Visualization ParaGraph and POLKA The IS collects runtime data in the form of time ordered trace records. These trace records are used to drive hard coded (ParaGraph) or user defined (POLKA) visualizations of system and program behavior. References: ....

Yan, Jerry C., S. R. Sarukkai, and P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit,", Software Practice and Experience, 25(4), April 1995, pp. 429--461.

....particular parallel programming model. INRIA SVMview: a Performance Tuning Tool for DSM based Parallel Computers 11 Several performance tuning tools have been designed in the past, but most of them focused mainly on performance tuning for message passing parallel codes like Paragraph [16] or AIMS [26]. None of these tools can provide graphical views adapted to our requirements (i.e. able to show the DSM activity with a relationship to the high level programming source code) Moreover, few of them are really available. Instead of modifying an existing tool, another approach consists in using a ....

J. Yan, S. Sarukkai, and P. Mehra. Performance measurement visualization and modeling of parallel and distributed programs using the AIMS toolkit. Software, Practice and Experience, 25(5):429--461, April 1995.

....the Recorder. We have chosen not to use the recording facilities found in TNF [4] although the technique is similar to our Recorder. The main reason is that TNF uses a circular buffer to store the recorded information and thus information may be overwritten if the buffer is too small. In [16] and [20] the authors stress the following issues: selective representation; integration between development time and run time information; high level and automated performance debugger; automated instrumentation of parallel programs; low overhead in monitoring program execution; and graphs and indices to ....

J. Yan, S. Sarukkai, and P. Mehra, "Performance Measurements, Visualization and Modelling of Parallel and Distributed Programs using the AIMS Toolkit," Software-Practice and Experience, 25(4):429-461, Apr. 1995.

....to its application specific workload characterization technique, which is distinctive for its short turn around time and less dependence on the low level system measurements. Similar workload characterization techniques are used in several simulation based performance modeling and prediction tools [2,26]. The objective of this modeling based evaluation is to choose among different options for the design of the adaptive controller. We compare the design alternatives for the controller on the basis of their maintainability of the application execution close to the desired operating range. Adaptive ....

....knowledge about the application domain. This type of workload characterization is increasingly becoming popular for performance prediction studies that use a simulation models, which is parameterized for a particular parallel or distributed system using only highlevel, coarse grained measurements [2,26]. In order to model the behavior of an application, we first divide it into smaller and manageable modules that may have complex patterns of interdependence. A particular module of an application can be a process or a thread that performs a series of high level functions distinguished by their ....

Yan, Jerry C., S. R. Sarukkai, and P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit,", Software Practice and Experience, 25(4), April 1995, pp. 429--461.

....even obtaining accurate trace information. Mohr, Brown, and Malony have developed TAU (Tuning and Analysis Utility) MoB94] which uses profiling and tracing information for parallel program development and analysis. AIMS (Automated Instrumentation and Monitoring System) developed by Yan et al. [YaS95] provides an automatic source code instrumentor, a runtime performance monitoring library, a trace animation and analysis toolkit, and a trace post processor to partially compensate for data collection intrusion. Of course, it is not entirely realistic to decouple task determination from ....

J.C. Yan, S.R. Sarukkai, and P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit," Software Practice and Experience (to appear), 1995.

....tool, a semi automated performance diagnosis system, and a database of program measurements. Similarly, use of modeling and simulation tools to support scalability analysis has been demonstrated by Yan, Sarukkai, and Mehra with the AIMS (Automated Instrumentation and Monitoring System) toolkit [26], a toolset for tuning and predicting the performance of large scale, message passing parallel and distributed scientific applications. Trace files from AIMS modeling tool can be read by its visualization and profiling tools to analyze simulated performance. Additionally, program execution traces ....

Yan, J., S. Sarukkai, and P. Mehra, "Performance Measurement, Visualization, and Modeling of Parallel and Distributed Programs using the AIMS Toolkit," Software--Practice and Experience, 25(4), April 1995, pp. 429-461.

....analyze dynamic execution characteristics. In the past five years, visualization technologies has been applied to manage the vast amount of performance data collected in multiprocessors. Many tools supporting performance visualization are available today, either in the public domain (e.g. AIMS [YSM95], Pablo [RO 91] Paradyn [MC 95] and ParaGraph [Hea93] or as part of a multiprocessor s system software (e.g. CXTRACE on Convex s SPP 1, ParAid on Intel s Paragon, MPP Apprentice on CRI s T3D, PRISM on TMC s CM5, and PV on IBM s SP2) Typically, performance data are displayed postmortem. These ....

....variety of analyses using fixed length data files the instrumentation, monitoring and trace postprocessing environment on which these approaches are tested is first described in the next section. 1. 3 Performance Tuning with AIMS The Automated Instrumentation and Monitoring System , or AIMS [YSM95], consists of a suite of software tools for measurement and analysis of parallel program performance. An instrumentor first parses the application s source code to insert instrumentation at selected points. Upon execution of the instrumented program (linked with a machine specific monitor ....

J. C. Yan, S. R. Sarukkai, and P. Mehra. "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit". Software Practice & Experience. April 1995. Vol. 25, No. 4, pages 429-461

....analyze dynamic execution characteristics. In the past five years, visualization technologies has been applied to manage the vast amount of performance data collected in multiprocessors. Many tools supporting performance visualization are available today, either in the public domain (e.g. AIMS [YSM95], Pablo [RO 91] Paradyn [MC 95] and ParaGraph [Hea93] or as part of a multiprocessor s system software (e.g. CXTRACE on Convex s SPP 1, ParAid on Intel s Paragon, MPP Apprentice on CRI s T3D, PRISM on TMC s CM5, and PV on IBM s SP2) Typically, performance data are displayed postmortem. These ....

....variety of analyses using fixed length data files the instrumentation, monitoring and trace postprocessing environment on which these approaches are tested is first described in the next section. 1. 3 Performance Tuning with AIMS The Automated Instrumentation and Monitoring System , or AIMS [YSM95], consists of a suite of software tools for measurement and analysis of parallel program performance. An instrumentor first parses the application s source code to insert instrumentation at selected points. Upon execution of the instrumented program (linked with a machine specific monitor ....

J. C. Yan, S. R. Sarukkai, and P. Mehra. "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit". Software Practice & Experience. April 1995. Vol. 25, No. 4, pages 429-461

....of the environment that do not change are preserved without the need to model them. This allows us to incrementally target parameters and study the effect of such changes. Secondly, projected performance in the new environment can be studied with a whole suite of trace analysis tools in AIMS [YSM94] that help pin point source code locations that could be potential bottlenecks, as well as their possible causes [SYG94] This technique also has some limitations: it cannot be used to predict the performance of a different problem size or a multiprocessor with different number of nodes. Such ....

....a different problem size or a multiprocessor with different number of nodes. Such predictions require a more comprehensive understanding of program structure; preliminary studies have been carried out based on simulation as well as analytical techniques dealt with the performance prediction tools [SaM94, YSM94]. The tool discussed here simply provides a convenient means of gauging the impact of changes in the execution environment for one instance of the problem size and number of processing nodes. 2 Analysis Methodology We first automatically insert instrumentation in the program to generate ....

[Article contains additional citation context not shown here]

Jerry Yan, Sekhar R. Sarukkai and Pankaj Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs," to appear in the Journal of Software Practice and Experience.

....ParAide for Intel s Paragon, MPP Apprentice for CRI s T 3D, PRISM for TMC s CM 5, PV for IBM s SP2 and CXTrace for Convex s SPP1. Many public domain tools have also emerged from various national laboratories and universities including UIUC s Pablo [RO 91] ORNL s XPVM [Kohl95] and NASA s AIMS [Yan94, YSM94]. A few observations can be made: Tools provided by the vendors only run on their own machines (as expected) Compounded with unreliable system software, a lot of these sophisticated performance tools are glanced over by many users who prefer gprof (which is supported in one form or another by ....

J. C. Yan, Sekhar R. Sarukkai and Pankaj Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs," Accepted to appear in the Journal of Software Practice and Experience.

No context found.

J. C. Yan, S. R. Sarukkai, and P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit," Software Practice & Experience, Vol. 25, No. 4, pp. 429-61.

No context found.

J. C. Yan, S. R. Sarukkai, and P. Mehra. "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit," Software Practice & Experience. April 1995. Vol. 25, No. 4, pages 429-461.

No context found.

Yan,J.,Sarukkai,S.,Mehra,P.:"Performancemeasurement, visualizationandmodelingofparallelanddistributedprograms usingtheAimstoolkit",SoftwarePracticeandExperience25-4 (1995),pp.429---461.

No context found.

Jerry Yan, Sekhar Sarukkai, and Pankaj Mehra, `Performance measurement, visualization and modeling of parallel and distributed programs using the aims toolkit', Software--Practice and Experience, 25(4), 429--461 (1995).

No context found.

J. Yan, S. Sarukkai, P. Mehra, "Performance Measurement, Visualization and Modeling of Parallel and Distributed Programs using the AIMS Toolkit", Software - Practice and Experience, Vol. 25, No. 4, (April 1995), pp. 429 - 461.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC