Optimization and solvers
What is SCIP Optimization Suite? An introduction to Zuse Institute Berlin's mixed-integer optimization solver suite
The SCIP Optimization Suite is an open-source collection of optimization software built around SCIP, a solver for mixed-integer linear and nonlinear programs. SCIP also works as a framework for constraint integer programming and branch-cut-and-price. It is made by Zuse Institute Berlin.
- Tool
- SCIP Optimization Suite
- Made by
- Zuse Institute Berlin
- Official site
- scipopt.org
- Category
- Optimization and solvers
Who makes SCIP Optimization Suite
The SCIP Optimization Suite is made by Zuse Institute Berlin. The official website is linked below.
A Getting started page is provided there for newcomers, and the SCIP OptSuite blog covers topics such as concurrent SCIP.
Official website: scipopt.org
What SCIP Optimization Suite is for
SCIP is used to solve mixed-integer linear programs (MIPs) and mixed-integer nonlinear programs (MINLPs). It gives users control of the solution process rather than treating the solver as a sealed box.
It is also a framework. Beyond solving standard problem classes, it supports constraint integer programming and branch-cut-and-price, which lets users build specialised solution methods on top of it.
Who SCIP Optimization Suite is for
The vendor does not state a specific audience. The emphasis on user control and on exposing detailed solver information suggests it suits people who want to inspect and adjust how a solver works, such as researchers and developers of optimization methods.
A dockerized SCIP container for teaching has been described in a blog post, which points to use in instruction as well.
Notable capabilities
The Suite combines a core solver with several companion tools and a set of default plugins.
- MIP and MINLP solving: SCIP is an open-source solver for mixed-integer linear and nonlinear programs.
- User control: SCIP gives users control of the solution process and exposes detailed solver information.
- Default plugins: SCIP ships with a set of default plugins for solving MIPs and MINLPs.
- Framework use: SCIP is also a framework for constraint integer programming and branch-cut-and-price.
- Bundled components: the Suite bundles SCIP with GCG, SoPlex, ZIMPL, PaPILO and UG.
- Permissive licensing: SCIP moved to the Apache 2.0 License.
- Public repositories: the main development branches of SCIP, SoPlex and PaPILO are mirrored publicly on GitHub, and symmetry detection uses a fork of bliss hosted on GitHub.
How SCIP Optimization Suite fits optimization and solvers
In the category of optimization and solvers, the SCIP Optimization Suite covers both ready-to-use solving and customisable framework work. Its scope includes nonlinear mixed-integer problems as well as linear ones, and it comes with related tools for modeling, linear programming, presolving and parallelisation in the bundled components.
For Reasonelm's wider niche of decision intelligence, decision support and decision analytics, a solver like this is a back-end component. When a decision involves discrete choices and constraints, such as selecting options or sequencing activities, the problem can be expressed as a mixed-integer model and passed to a solver. The visibility SCIP offers into the solution process can help analysts understand how a recommended answer was reached.
Frequently asked questions
What is the SCIP Optimization Suite?
It is an open-source solver and framework for mixed-integer linear and nonlinear programming and constraint integer programming. It bundles SCIP with GCG, SoPlex, ZIMPL, PaPILO and UG.
Who develops the SCIP Optimization Suite?
It comes from Zuse Institute Berlin.
What licence does SCIP use?
SCIP moved to the Apache 2.0 License.
Does SCIP support branch-cut-and-price?
Yes. SCIP is also a framework for constraint integer programming and branch-cut-and-price.
Where can newcomers start with SCIP?
A Getting started page is provided for newcomers, and the SCIP OptSuite blog covers further topics such as concurrent SCIP.
Reasonelm is independent and is not affiliated with Zuse Institute Berlin. Product names belong to their owners. Features change; confirm details on the official website.