Colloq: Speaker Institution:
Mississippi State University
Colloq: Date and Time:
Wed, 2015-04-01 10:00
Building 5100, Room 262 (Boardroom)
Colloq: Host Email:
Recent developments in the field of parallel and distributed computing has led to a proliferation of solving large and computationally intensive mathematical, science, or engineering problems, that consist of several parallelizable parts and several non-parallelizable (sequential) parts. In a parallel and distributed computing environment, the performance goal is to optimize the execution of parallelizable parts of an application on concurrent processors. This requires efficient application scheduling and resource allocation for mapping applications to a set of suitable parallel processors such that the overall performance goal is achieved. However, such computational environments are often prone to unpredictable variations in application (problem and algorithm) and system characteristics. Therefore, a robustness study is required to guarantee a desired level of performance. Given an initial workload, a mapping of applications to resources is considered to be robust if that mapping optimizes execution performance and guarantees a desired level of performance in the presence of unpredictable perturbations at runtime. In my research, a stochastic process algebra, Performance Evaluation Process Algebra (PEPA), is used for obtaining resource allocations via a numerical analysis of performance modeling of the parallel execution of applications on parallel computing resources. The PEPA performance model is translated into an underlying mathematical Markov chain model for obtaining performance measures. Further, a robustness analysis of the allocation techniques is performed for finding a robust mapping from a set of initial mapping schemes. The numerical analysis of the performance models have confirmed similarity with the simulation results of earlier research available in existing literature. When compared to direct experiments and simulations, numerical models and the corresponding analysis are easier to reproduce, do not incur any setup or installation costs, do not impose any prerequisites for learning a simulation framework, and are not limited by the complexity of the underlying infrastructure or simulation libraries.
Colloq: Speaker Bio:
Srishti Srivastava is a PhD candidate at the Department of Computer Science and Engineering at Mississippi State University since August 2010. Her research interests include dynamic load balancing, high performance computing, performance modeling, optimization, and prediction, robustness analysis of resource allocations, and autonomic computing. Presently, she is also a graduate research assistant at the Center for Autonomic Computing at Mississippi State University, which is also one of the four National Science Foundation sites for autonomic computing. Srishti has authored and co-authored a number of articles published in renowned IEEE and ACM conferences, journals, and book chapters. Srishti has also been selected to be a part of a delegation of 200 young scientists from around the world to attend the Heidelberg Laureate Forum (to meet laureates in Math and Computer Science) at Heidelberg, Germany, October 2014. She is the member of the IEEE computer society, ACM, Society for Industrial and Applied Mathematics (SIAM), Computing Research Association (CRA, CRA-W), Anita Borg Institute Grace Hopper Celebration (ABI-GHC), and an honor society of Upsilon Pi Epsilon (UPE).