• PASTIS is a distributed-memory parallel many-to-many protein sequence aligner that uses sparse matrices.

  • Sphynx is a distributed-memory spectral graph partitioner that can take advantage of accelerators such as GPUs.

  • Ripples (and cuRipples) is a software framework to study the Influence Maximization problem.

  • Zoltan2 is a package in the Trilinos scientific computing toolkit/library that provides partitioning, load-balancing, and other combinatorial scientific computing features. It is fully parallel and has been used on supercomputers with hundreds of thousands of cores. It is the successor to the well-known Zoltan package, but has better support for new architectures (such as KNL, GPU) and tighter integration into Trilinos.

  • The Combinatorial BLAS (CombBLAS) is an extensible distributed-memory parallel graph library offering a small but powerful set of linear algebra primitives specifically targeting graph analytics. It powers the HipMCL protein clustering software of the ExaBiome project and provides a foundation for the GraphBLAS standardization effort.

  • Heavy-weight perfect matching (HWPM) code helps distributed sparse solvers such as SuperLU_dist and STRUMPACK to scale to larger concurrencies by eliminating serial bottlenecks. HWPM code is part of CombBLAS. Read more about HWPM here.

  • KokkosKernels is a library for shared-memory parallel graph and (sparse & dense) matrix kernels. Although KokkosKernels is primarily funded and developed by SNL/ATDM, ExaGraph contributes to the graph algorithms, such as graph coloring.