A framework to develop symbolic performance models of parallel applications Export

@article{citeulike:4294357, abstract = {Performance and workload modeling has numerous uses at every stage of the high-end computing lifecycle: design, integration, procurement, installation and tuning. Despite the tremendous usefulness of performance models, their construction remains largely a manual, complex, and time-consuming exercise. We propose a new approach to the model construction, called modeling assertions (MA), which borrows advantages from both the empirical and analytical modeling techniques. This strategy has many advantages over traditional methods: incremental construction of realistic performance models, straightforward model validation against empirical data, and intuitive error bounding on individual model terms. We demonstrate this new technique on the NAS parallel CG and SP benchmarks by constructing high fidelity models for the floating-point operation cost, memory requirements, and MPI message volume. These models are driven by a small number of key input parameters thereby allowing efficient design space exploration of future problem sizes and architectures.}, address = {Los Alamitos, CA, USA}, author = {Alam, S. R. and Vetter, J. S.}, citeulike-article-id = {4294357}, citeulike-linkout-0 = {http://doi.ieeecomputersociety.org/10.1109/IPDPS.2006.1639625}, citeulike-linkout-1 = {http://dx.doi.org/10.1109/IPDPS.2006.1639625}, citeulike-linkout-2 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1639625}, doi = {10.1109/IPDPS.2006.1639625}, journal = {Parallel and Distributed Processing Symposium, International}, keywords = {parallel\_algorithms, parallel\_programming, performance\_modelling}, pages = {368+}, posted-at = {2009-04-09 13:08:17}, priority = {2}, publisher = {IEEE Computer Society}, title = {A framework to develop symbolic performance models of parallel applications}, url = {http://dx.doi.org/10.1109/IPDPS.2006.1639625}, volume = {0}, year = {2006} }

TY - JOUR ID - citeulike:4294357 L3 - citeulike-article-id:4294357 N2 - Performance and workload modeling has numerous uses at every stage of the high-end computing lifecycle: design, integration, procurement, installation and tuning. Despite the tremendous usefulness of performance models, their construction remains largely a manual, complex, and time-consuming exercise. We propose a new approach to the model construction, called modeling assertions (MA), which borrows advantages from both the empirical and analytical modeling techniques. This strategy has many advantages over traditional methods: incremental construction of realistic performance models, straightforward model validation against empirical data, and intuitive error bounding on individual model terms. We demonstrate this new technique on the NAS parallel CG and SP benchmarks by constructing high fidelity models for the floating-point operation cost, memory requirements, and MPI message volume. These models are driven by a small number of key input parameters thereby allowing efficient design space exploration of future problem sizes and architectures. JF - Parallel and Distributed Processing Symposium, International AD - Los Alamitos, CA, USA TI - A framework to develop symbolic performance models of parallel applications PB - IEEE Computer Society VL - 0 SP - 368 KW - parallel_algorithms KW - parallel_programming KW - performance_modelling AU - Alam, SR AU - Vetter, JS PY - 2006/// UR - http://dx.doi.org/10.1109/IPDPS.2006.1639625 ER -