We are the Combinatorial Scientific Computing Lab at the University of California, Santa Barbara.

We focus on problems involving computations on large graphs. As a result, our research is often conducted at the intersection of several fields, including numerical linear algebra, parallel computing, and scientific computing.

People at the CSC Lab

Dr. John Gilbert

Current Members:
Kevin Deweese, Veronika Strnadova

Past Members
Aydin Buluc, Viral B. Shah, Vikram Aggarwal, Imran Patel, Varad Deshmukh, and Adam Lugowski

For more information about past and current lab members, please refer to our People page.


Current Projects

  • Knowledge Discovery Toolkit (KDT): A toolkit which provides domain experts with a simple interface to analyze very large graphs quickly and effectively without requiring knowledge of the underlying graph representation or algorithms.
  • The Combinatorial BLAS (CombBLAS): High-performance parallel primitives for graph analysis and mining. The primitives are based on sparse linear algebra and implemented in MPI/C++ for distributed memory parallelism. 

Past Projects

  •  High Productivity Computer Systems: Petascale computing with an emphasis on programmer productivity
  • Star-P: A tool for interactive numerical and combinatorial computing.
  • Circuitscape: Using circuit theory to design connected landscapes.