- Paper on Fast Linear Algebra-Based Triangle Counting with KokkosKernels, got selected as the "Champion" in the IEEE/Amazon/DARPA Graph Challenge (with Michael Wolf, Mehmet Deveci, Jon Berry, and Si Hammond).
- Paper on Ensemble grouping strategies for embedded stochastic collocation methods applied to anisotropic diffusion problems got accepted to SIAM/ASA Journal on Uncertainty Quantification (with Marta D'Elia, Eric Phipps, Jonathan Hu, Carter Edwards).
- Paper on Designing Vector-Friendly Compact BLAS and LAPACK Kernels got accepted to Supercomputing 2017 (with Kyungjoo Kim, Tim Costa, Mehmet Deveci, Andrew Bradley, Simon Hammond, Murat Guney, Sarah Knepper, and Shane Story).
- Paper on Basker: Parallel Sparse LU Factorization utilizing Hierarchical Parallelism and Data Layouts, with Joshua Booth, Nathan Ellingwood and Heidi Thornquist got accepted in Parallel Computing.
- Embedded Ensemble Propagation for Improving Performance, Portability, and Scalability of Uncertainty Quantification on Emerging Computational Architectures, with Eric Phipps and others just appeared in SIAM Journal on Scientific Computing.
- Paper on Partitioning Trillion-edge graphs in minutes got accepted to IPDPS 2017 (with George Slota and Kamesh Madduri).
- Paper on Performance-Portable Sparse Matrix-Matrix Multiplication for Many-Core Architectures got accepted to Accelerators and Hybrid Exascale Systems workshop at IPDPS 2017 (with Mehmet Deveci and Christian Trott).
- Best paper award in the Seventh International Workshop on Accelerator and Hybrid Exascale Systems (AsHES) workshop at IPDPS 2017.
- Paper on Order or Shuffle: Empirically Evaluating Vertex Order Impact on Parallel Graph Computations got accepted to Graph Algorithms Building Blocks workshop at IPDPS 2017 (with George Slota and Kamesh Madduri).
- Complex Network Partitioning Using Label Propagation, with George Slota and Kamesh Madduri just appeared in SIAM Journal on Scientific Computing.
- The new github organization HPCGraphAnalysis has the codes for
- Multistep - the implementation of multithreaded (strongly) connected components algorithm
- HPCGraph - a library for different graph computations on HPC systems
- PuLP - a shared-memory and distributed-memory graph partitioning tool using label propagation
- We just posted our report on Partitioning Trillion-edge graphs in minutes in arXiv (with George Slota and Kamesh Madduri).
- A talk I gave in the Sixth International Workshop on Accelerators and Hybrid Exascale Systems (AsHES) in IPDPS 2016 can be accessed here.
- Long time graduate student intern George Slota moved to RPI as an assistant professor. Congrats, George !
- A talk I gave in the fourteenth copper mountain conference on iterative methods can be accessed here.
- We submitted an invited journal article on survey of direct methods for sparse linear systems to Acta Numerica. The Tech Report version of the survey of direct methods is also available.
(505) 844-7181 (Phone)
Mailing address (USPS)
Sandia National Laboratories
P.O. Box 5800, MS 1320
Albuquerque, NM 87185-1320
Sandia National Laboratories
1515 Eubank SE
Albuquerque, NM 87123