DrRavi Manumachu

Heterogeneous Parallel Computing
______________________________
Functional performance and energy models of computation, model-based methods and data partitioning algorithms for optimization of applications for performance and energy on heterogeneous computing platforms, heterogeneous parallel algorithms, programming systems and message-passing libraries for heterogeneous computing platforms, and linear algebra libraries for heterogeneous computing platforms.

Energy-efficient Computing
___________________________
System-level and application-level energy efficiency techniques, theory of energy predictive models, energy predictive models, model-based methods for application-level energy optimization, and software component-level energy decomposition of hybrid parallel applications.

Parallel and Distributed Machine Learning
_________________________________________
Performance optimization of stochastic gradient descent in parallel and distributed deep learning.