Sparse linear algebra and parallel algorithms

01 / Sparse computation

Sparse matrices,
tensors, and solvers

At NVIDIA, I primarily work on our next-generation iterative solvers (cuISS), but I also frequently collaborate with cuDSS and cuSPARSE teams.

My broader research includes parallel sparse matrix multiplication, sparse matrix-vector multiplication, sparse tensors, distributed data structures, and communication-avoiding algorithms. A recurring goal is to reduce data movement while exposing parallelism across cores, GPUs, and distributed-memory systems.

These methods support scientific computing, graph analysis, and parallel algorithms for machine learning, including communication-efficient graph neural network training.

02 / Graph algorithms

Linear algebra as
a language for graphs

03 / Computational biology

From sparse computation
to protein families

I conceived and led the design of HipMCL, a scalable parallel implementation of Markov clustering, and PASTIS, a distributed-memory protein sequence similarity tool based on sparse matrix operations.

I also contributed to the HipMer/MetaHipMer genome assembly effort and to protein-family identification and genome assembly at supercomputing scale within ExaBiome.

Research leadership

Collaborative science

Sparsitute

I was the founding lead PI and director of Sparsitute from its founding in September 2022 until November 2025. The DOE mathematical center spans six institutions and more than 30 members. I set the technical agenda and coordinated research in sparse computation and software.

I served as LBNL PI for ExaGraph and led research programs in scalable graph learning, randomized sparse algorithms, and energy-efficient parallel graph and data mining algorithms.

Professional service

Conference and workshop organization

Workshop organization

• 2025 — Chair: Workshop on Network Algorithms, Analysis, and Learning for Science.
• 2023 — Co-chair: JRG70: Workshop Celebrating the Scientific Contributions of John R. Gilbert.
• 2022 — Co-chair: First ACDA Workshop in Aussois.
• 2020–2021 — Co-chair: Randomized Algorithms for Scientific Computing (RASC), a DOE-sponsored workshop.
• 2019 — Co-chair: ASCR Applied Mathematics Principal Investigators Meeting.
• 2019 — Organizing committee: First Applied Math Visioning Workshop: The Future of Machine Learning and Data Analytics Across the Department of Energy.

Conference and committee leadership

• 2025 — Area Chair: ACM Symposium on Parallelism in Algorithms and Architectures (SPAA).
• 2023 — Organizing committee: Workshop on Sparse Tensor Computations.
• 2020 — Organizing committee: SIAM Conference on Parallel Processing for Scientific Computing.
• 2018 — Algorithms Track Chair: ACM/IEEE SC.
• 2018 — Publications Chair: SIAM Workshop on Combinatorial Scientific Computing (CSC).
• 2017 — Applications Track Vice Chair: ACM/IEEE SC.
• 2017 — Program Co-chair: IEEE Graph Algorithm Building Blocks (GABB) workshop at IPDPS.
• Steering committees: GABB at IPDPS (2014–2018); Graphs, Architectures, Programming, and Learning (GrAPL) at IPDPS (joined 2019).

Minisymposia organized or co-organized

• 2024: Large Scale Graph Analytics — SIAM Conference on Parallel Processing for Scientific Computing.
• 2023: Sparse Computations in Science and Engineering — SIAM Conference on Computational Science and Engineering.
• 2021: GraphBLAS: Tools, Algorithms, and Applications — SIAM Conference on Computational Science and Engineering.
• 2019: Combinatorial Scientific Computing — International Congress on Industrial and Applied Mathematics (ICIAM).
• 2016: GraphBLAS: Graph Algorithms in the Language of Linear Algebra — SIAM Annual Meeting.
• 2016: Combinatorial Scientific Computing — SIAM Conference on Parallel Processing for Scientific Computing.
• 2014: Graph Analysis for Scientific Discovery — SIAM Conference on Parallel Processing for Scientific Computing.
• 2013: Minimizing Communication in Scientific Computing — SIAM Conference on Computational Science and Engineering.
• 2012: Reducing Communication in Linear Algebra — SIAM Conference on Applied Linear Algebra.
• 2011: Optimization Challenges in the Energy Sector — SIAM Conference on Optimization.

Tutorials and community sessions

• 2021 — Tutorial organization: Building Graph Algorithms with the GraphBLAS, a minitutorial at SIAM CSE; A Hands-On Introduction to GraphBLAS at the International Conference on Supercomputing (ICS).
• 2021 — Guided Affinity Group Leader: Broader Engagement Program, SIAM CSE.
• 2017–2018 — Co-organizer: HPC Graph Toolkits and GraphBLAS Forum birds-of-a-feather session, SC.
• 2014–2015 — Session leader: GraphBLAS Forum birds-of-a-feather session, HPEC.

Editorial and society service

• Associate Editor, SIAM Journal on Scientific Computing; founding Associate Editor, ACM Transactions on Parallel Computing (2013–2020).
• Guest Editor, ACM Transactions on Parallel Computing, special issue for SPAA 2018 (2021); Parallel Computing, special issue on Graph Analysis for Scientific Discovery (2015).
• SIAM Membership Committee (2022–2024); George Pólya Prize for Mathematical Exposition Selection Committee (2025); SIAG/Linear Algebra Early Career Prize Committee for the 2027 award.