| Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding of the 9th IEEE/ACM HighPerformance Networking and Computing Conference (SC'2001), Denver, USA, November 2001. |
....undertaken to provide a test suite to check the correctness of performance analysis tools. Such a suite would immediately benefit all developers of automatic performance analysis tools. Currently, members of the APART working group are developing prototypes of automatic performance analysis tools [1, 2, 3, 4, 5, 6]. They also have solid experience in collaborative work in the group. This makes APART a natural forum for the design of such a test suite. In this report we discuss the initial design of a test suite, which we from now on will simply term ATS (for APART Test Suite) for checking the correctness ....
....The ASL report also includes a catalog of typical performance properties for MPI, OpenMP and HPF programs. For a hybrid MPI OpenMP programming style, especially with the Hitachi SR 8000 in mind, this catalog was extended in [8] Similar performance properties can be found formalized in [1, 2] and [4, 5]. These typical properties can form the basis for the ATS framework which allows to construct synthetic positive and negative test programs. The initial design for this framework is presented in Chapter 3. Finally, is has to be ensured that a performance analysis tool not only works for carefully ....
H.-L. Truong, T. Fahringer, G. Madsen, A. D. Malony, H. Moritsch, S. Shende. n Using SCALEA for Performance Analysis of Distributed and Parallel Programs. SC 2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding of the 9th IEEE/ACM HighPerformance Networking and Computing Conference (SC'2001), Denver, USA, November 2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding SC'2001, Denver, USA, November 2001. IEEE/ACM.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding of the 9th IEEE/ACM HighPerformance Networking and Computing Conference (SC'2001), Denver, USA, November 2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding SC'2001.
No context found.
H. Truong, T. Fahringer, G. Madsen, A. Malony, H. Moritsch and S. Shende. On using SCALEA for Performance Analysis of Distributed and Parallel Programs. Supercomputing 2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding SC'2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding SC'2001, Denver, USA, November 2001. IEEE/ACM.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding of the 9th IEEE/ACM HighPerformance Networking and Computing Conference (SC'2001), Denver, USA, November 2001.
No context found.
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding of the 9th IEEE/ACM HighPerformance Networking and Computing Conference (SC'2001), Denver, USA, November 2001.
....formalizes trace files by using an event class. This paper does not contain any property which incorporates events, which may change in future work. 5 Note that due to advanced monitoring and profiling technologies (e.g. dynamic profiling [14] hardware profiling [5] and source code profiling [19, 15]) there is basically no barrier to obtain experiment related data for arbitrary parallel and distributed programs. 4.2 Filters and Statistics for Experiment related Data It order to provide flexible mechanisms to describe performance problems we commonly require filters (used to restrict the ....
....outlined in Figure 1 that uses JavaPSL to determine performance properties for MPI, OpenMP, and mixed OpenMP MPI programs. Our current prototype tries to compute every performance property defined in this paper for user definable code regions based on the experiment related data available. SCALEA [19] is used to instrument and profile input programs. Moreover, profile data is automatically converted to the format defined in Section 4.1. In this section, we evaluate our prototype implementation by applying it to several applications executed on Gescher, an SMP cluster with 6 SMP nodes ....
Hong-Linh Truong, Thomas Fahringer, Georg Madsen, Allen D. Malony, Hans Moritsch, and Sameer Shende. On Using SCALEA for Performance Analysis of Distributed and Parallel Programs. In Proceeding SC'2001.
....we describe the choice of the cost functions for the MetaPL CodeBlocks corresponding to the code fundamental units. Fig. 3 illustrates the basic steps for building the cost functions. The first step is to instrument the parts of the application code that correspond to code blocks. The SCALEA tool [14] was used to instrument the code and gather the time spent in the application execution. Even if it is possible to automatically instrument the source code, we used a manual approach to define the code blocks and the SCALEA code regions . Then the tests to be executed were chosen. The tests are ....
H. Truong, T. Fahringer, G. Madsen, A. Malony, H. Moritsch and S. Shende. On using SCALEA for Performance Analysis of Distributed and Parallel Programs. Supercomputing 2001.
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