Frontline Systems adds Hexaly to Analytic Solver for Excel
Frontline Systems has integrated Hexaly into its latest Analytic Solver release, giving Excel users a single solver that handles a wide range of models, from linear formulations to spreadsheets full of IF and LOOKUP formulas. Tested by Frontline on hundreds of models, the new engine is described as the company’s most versatile solver to date.
The challenge: optimization for people who think in Excel
Frontline Systems makes optimization accessible to Excel users. Its Analytic Solver product family lets analysts, engineers, and business users build optimization and simulation models directly in Excel, without having to learn the mathematical taxonomy of linear, convex quadratic, smooth nonlinear, and non-smooth problems.
But real-world spreadsheets rarely stay tidy. They are full of conditional logic, table lookups, and functions that break the assumptions most classical solvers rely on. Frontline wanted an engine that could stay effective across that whole range, from textbook mixed-integer models to arbitrary, formula-heavy ones.
The solution: Hexaly Optimizer inside Analytic Solver
With the release of Analytic Solver V2026 Q4, Frontline Systems introduced a new multi-purpose solver engine based on Hexaly Optimizer. After testing it on hundreds of models, from linear mixed-integer to smooth nonlinear to non-smooth problems, Frontline describes Hexaly as its most versatile solver to date, with excellent performance across the full spectrum of models.
The Hexaly engine is available in two editions:
- Standard Hexaly Solver: included at no extra cost in Analytic Solver Optimization and Analytic Solver Comprehensive, with the same size limits as Frontline’s other standard solvers (1,000 decision variables for nonlinear and non-smooth models, 8,000 for linear models).
- Large-Scale Hexaly Solver: an optional add-on with no fixed limits beyond available memory and compute time. Frontline’s tests show it is effective on very large linear, quadratic, nonlinear and reasonably structured non-smooth models.
Why Hexaly fits: a different approach to optimization
Hexaly’s search methods differ from the branch-and-bound approaches used by other solver engines and from gradient-based methods. As Frontline notes, this is why Hexaly can often perform remarkably well across very different model types, from classical linear mixed-integer formulations to models built on IF, CHOOSE, LOOKUP and similar Excel functions.
A powerful combination: Hexaly and the PSI Interpreter
Hexaly is most effective when a model’s calculations are exposed in a form the optimizer can understand, rather than as a black box. This is where Frontline’s Polymorphic Spreadsheet Interpreter (PSI) comes in. PSI understands how to calculate nearly all of Excel’s roughly 500 built-in functions, including functions that “break the rules” of linear and smooth models. For this release, Frontline extended PSI to convert more Excel models into the form Hexaly works best with.
For users who know Excel better than optimization theory, this is a new kind of ease of use. It remains good practice to express a model in linear or smooth form when that reflects reality, but it is now much easier to get high-quality solutions, with strong performance, when relationships involve conditionals or table lookups.
What’s next
Frontline Systems continues to evolve Analytic Solver and its cloud platform RASON, with further releases planned. Users can try the new engine by installing the latest SolverSetup for Analytic Solver Desktop, or by using Analytic Solver Cloud, which is always up to date.