• Recent research interest and expertise – Combinatorial scientific computing novel partitioning models, algorithms and software utilities for development of parallel iterative methods for linear-system solutions – Optimizing latency-centric communication metrics for petascaling sparse solvers – Partitioning irregular domains for large- scale parallel processing – Locality aware scheduling of irregular applications on Many Core architectures – Partitioning models for scaling 1D-, 2D- and 3D-parallel sparse matrix-matrix multiply – Partitioning large scale social networks and graph databases – Parallel graph analytics kernels for big data applications Speedup curves of Conjugate Gradientsolver for different methods on a Cray and BlueGene/Q machine (kkt-power matrix: 2 million rows, 12 million nonzeros) – HPC for Machine Learning and ML for HPC – Partitioning methods for scalable sparse Tensor decomposition – Scaling parallel stochastic gradient descent algorithms for ML – Fast and efficient model parallelism for Deep CNNs – Task leader in FP7 / Horizon2020 PRACE projects: 1IP, 2IP, 3IP, 4IP, 5IP |