SparSol - high-performance, sparse linear solver

Screen shot of SparSol

  • High-performance, iterative solver for very large systems of sparse linear equations – easily handles millions of unknowns
  • Solves both symmetrical and unsymmetrical systems including systems with block structures - excels at solution of very ill-conditioned systems
  • Adaptive convergence schemes guarantee fast convergence
  • Incorporates a large collection of custom, problem-specific algorithms with numerous widely-used, public domain algorithms
  • Highly-efficient and scalable parallel algorithms with support for both multi-core (SMP) and distributive (MPI) architectures
  • Quickly integrates into existing data analysis applications - supported on Windows, Linux and Unix


SparSol is a library of highly-efficient algorithms intended for the preconditioned iterative solution of large sparse linear algebraic systems of equations with real coefficients.

Sparse linear systems often arise when numerically solving partial differential equations that are common in scientific and engineering applications. Virtual prototyping applications, like those used in the aerospace, automotive and semiconductor industries, may handle millions of these equations when simulating complex products. Power grid simulation, oil and gas reservoir modeling and financial engineering are examples of large-scale modeling applications that struggle with sparse linear systems. The ability to solve these systems quickly and efficiently can have a direct and significant impact on productivity and profitability for many companies.

SparSol has years of proven results in some of the most computationally extreme, real-world applications.  Originally developed for ExxonMobil Upstream Research Company's oil reservoir modeling application, SparSol is now available worldwide from NeurOK Software to accelerate the most demanding analytical applications.  SparSol’s algorithms can quickly be integrated into existing applications through a rich set of application programming interfaces (APIs), delivering dramatic improvements in performance over existing sparse linear solvers.

Learn More