16 resources
There's a class of industrial optimization problems that sits just out of reach of classical solvers. The physics-based simulation model exists and is accurate — but it's a numerical black box.
Update (July 2026): The article has been extended and includes now result from SCIP Concurrent solver. SCIP-Conc is a parallel solver for mixed-integer programming, built on SCIP and has been around for some time. Recent activities for …
A hands-on comparison of ReLU modeling in GAMSPy, MIP vs NLP, for adversarial verification on realistic networks. The workflow, code snippets, and trade-off between speed and guarantees are all explained in details.
Presentation slides covering the evolution of CONOPT, recent performance and usability improvements, new APIs, and derivative-based nonlinear optimization workflows.
For a long time, GPUs were essential for various AI applications and other high-performance compute applications but had limited impact on mathematical optimization. That’s changing.
To support their carbon neutrality ambitions, the global energy company TotalEnergies developed a complex optimization model for Carbon Capture and Storage (CCS). This highly sophisticated model, capable of managing the complex CO2 …
The Austrian Power Grid (APG) partnered with our consulting team to optimize its electricity transmission model. As a result, model-solving time was slashed by over 60%, memory usage dropped by 80%, and various infeasibilities were …
Learn how GAMSPy revolutionizes mathematical optimization by seamlessly integrating the power of Python's expressive language with the efficiency of GAMS, offering a streamlined and high-performance solution for complex modeling challenges.
MINLP solvers typically have a rooting in academia. This article highlights the historic performance gains of two of them, BARON and SCIP.
A recent webinar by Evangelos Panos demonstrates how to tune CPLEX options to get more bang for your buck when running large problems such as the TIMES ESM framework.
With GAMS 37, we now support the generation of model instances with more than 2³¹ non-zeros. We expect huge models to become more important in the future.
Stadtwerke München have recently lifted their computationally expensive GAMS model into the cloud. This white paper gives a high level overview of the techniques used and the benefits of a cloud deployment over a traditional on-premise …
At the United States Military Academy in West Point, all students must graduate in four years. Each student's daily activities are a carefully regimented balance of academic, military, and physical requirements. The ~4,500 enrolled cadets …
Optimization problems occasionally yield unbounded solutions. To find the cause one can modify the model and solve it to gain information. This is done through the imposition of “artificially” large bounds. Linear programming solvers …
Learn how to diagnose and correct numerical scaling problems in GAMS models using scaling attributes, GAMSCHK, CONVERT, and solver diagnostics.
Learn how to diagnose infeasible GAMS models by adding artificial variables, analyzing distorted marginals, and isolating conflicting equations and variable bounds.