GAMS

Great to visit University of Zagreb / Sveučilište u Zagrebu this week for the Third General Meeting of ROAR-NET. 🇭🇷 Our very own Hamdi Burak U. took the stage to talk about a challenge many of us face in optimization: Finding the right balance between heuristic methods and exact solvers. While heuristics are often the go-to for robust performance in the field, complex problems still demand the precision of an exact mathematical approach. Burak’s presentation focused on how we can bridge that gap using #GAMSPy. He shared some specific workflows on: 🔸Integrating heuristics as standalone components within a larger GAMS framework. 🔸How the GAMS approach to modeling compares with emerging ROAR-NET methodologies. 🔸Using Python to facilitate these hybrid workflows for better precision. It was a productive few days comparing notes with peers from Hexaly, ESTECO, Dots & Lines Ltd. and the broader research community. A big thanks to ROAR-NET COST Action for the hospitality! Get started with GAMSPy 👉 https://lnkd.in/dpQKUMnY 📸 credit to: Barbara Blečić #GAMS #GAMSPy #Optimization #ROARNET #OperationsResearch #MathematicalModeling #Python

We've updated our blog on GPU-Accelerated Optimization with GAMS and NVIDIA cuOpt to cover the February 2026 release of the GAMS/cuOpt solver link (v0.0.5d) for NVIDIA cuOpt 26.02. This comprehensive update introduces major architectural and algorithmic enhancements for both Large-Scale LPs and Mixed-Integer Programming (MIP): 🔶New PSLP Presolver: Architected specifically for very large-scale LPs, this robust, dual-preserving presolve replaces the PaPILO presolver to prevent memory exhaustion and time-outs on massive instances. 🔶 Expanded Hardware Support: The solver link now supports arm64 CPU architectures alongside x86_64. 🔶Advanced MIP Capabilities: We have exposed numerous new MIP-related options, including mip_trace for detailed solve insights and mip_start to utilize initial variable levels as a starting MIP solution. 🔶Initial Solutions: Support has been added for passing initial primal and dual solutions for both LP and MIP models. 🔶Solution Verification: For users leveraging first-order GPU methods like PDLP, we detail how to use the GAMS/Examiner tool to independently verify primal feasibility, dual feasibility, and true optimality. Read the updated technical breakdown and deployment guide here: https://lnkd.in/eZBkzCzW Explore the potential of cuOpt with GAMS and GAMSPy for your most challenging optimization tasks. To get started with cuOpt (or any other solver) on GAMS Engine SaaS, request a free test account by contacting sales@gams.com. #GAMS #NVIDIA #cuOpt #Optimization #OperationsResearch #MixedIntegerProgramming #LinearProgramming #HPC

We’re seeing some incredible work coming out of Abdullah Gül University lately! Shortly after seeing their peers tackle nurse scheduling, we’ve spotted another impressive project using GAMS & CPLEX, this time solving complex 3D loading challenges for Metal matris. The team (Rumeysa, Fatma, Zişan and Serra) took on the "Tetris" of industrial logistics: loading cylindrical bonel carcasses. By moving away from heuristics to a Mixed Integer Programming (MIP) approach, they achieved: 🔶 Efficiency Gains: Boosted space utilization from 89.6% to over 93%. 🔶 Layered Precision: Mastered complex orientation and stacking constraints. 🔶 Sustainability: Better packing means fewer trucks, lower costs, and a smaller carbon footprint. It is one thing to learn the theory of layer-based modeling; it’s another to deliver a 4% efficiency gain for a real-world partner like Hasçelik Kablo A.Ş. From healthcare shifts to 3D logistics, it’s a privilege to see how GAMS & CPLEX can take even the most 'messy' variables and turn them into an optimal reality. Excellent work to the team and once again to the mentors at AGU for fostering such practical, high-impact applications of mathematical modeling! #Logistics #SupplyChain #GAMS #Optimization #IndustrialEngineering #MIP #Sustainability

If you missed our 2025 Solver Review, last year marked a noticeable shift in how large-scale optimization performance is achieved. Beyond incremental solver improvements, 2025 introduced broader adoption of first-order methods, deeper hardware integration, and the first steps toward a modernized solver interface in GAMS. Three technical takeaways worth highlighting: 🔶 GPU-Enabled PDHG for Large LPs: Primal-Dual Hybrid Gradient (PDHG) algorithms became widely available across Gurobi, Xpress, COPT, and HiGHS. For extremely large LPs, PDHG, optionally accelerated on NVIDIA GPUs, offers a scalable alternative where traditional methods may struggle. 🔶 The GSI Transition Begins: With GAMS 51, Gurobi and Xpress solver links were migrated to the new GAMS Solver Interface (GSI). While still layered on top of GMO, GSI enables more efficient memory usage and lays the groundwork for tighter solver integration going forward. 🔶 SCIP and Lindo updates on the Horizon: The upcoming GAMS 53 release, currently available in beta, will bring SCIP 10 with CONOPT integration and Lindo API 16 with improved performance. 🔗 Catch up on the 2025–2026 Solver Roadmap: https://lnkd.in/dVNskF7a #MathematicalOptimization #OperationsResearch #GAMS #OptimizationSolvers Gurobi Optimization FICO #SCIP

In high-stakes numerical software, the greatest achievements are often the ones the user never feels because the system remains rock-solid while the engine is entirely rebuilt. Renke Kuhlmann and Stephen Maher just pulled off a massive architectural overhaul of CONOPT, merging a net change of +27,000 / -20,000 lines of code. The goal was simple but high-risk: Transition from Fortran to C without a "Big Bang" rewrite. Shifting to bind(c) ensures Fortran and C share the exact same memory layout. This lets us surgically refactor deep-seated solver functions into C without side effects, making the system much easier to extend. We’re maintaining the solver’s reliability while modernizing the architecture for the next few decades of optimization. Check out CONOPT for yourself 👉 https://conopt.gams.com/ #GAMS #SoftwareEngineering #Optimization #HighPerformanceComputing #CONOPT #Innovation #Fortran #CProgramming

Modeling is often limited by compute power. By using HPC resources from the Gauss Centre for Supercomputing, the UNSEEN project has moved from a few dozen scenarios to over 11,000. This jump in scale is a game-changer for energy policy. German Aerospace Center (DLR) Jülich Supercomputing Centre (JSC) Zuse Institute Berlin, Technische Universität Berlin, Federal Ministry for Economic Affairs and Energy, Forschungszentrum Jülich, University of Graz

🥇 1st Place:  Winner Spotlight: GAMSPy Student Competition: Nikhil Trivedi 🥇 Nikhil is in his final year at the University of Wisconsin-Madison, dual-majoring in Computer Science and Atmospheric & Oceanic Sciences. He used GAMSPy to solve a high-stakes infrastructure problem: where to place weather radars for the best possible coverage. Nikhil’s model optimizes radar placement across Texas, accounting for tricky variables like the Earth’s curvature and terrain blockage. By maximizing high-resolution data for millions of people, his work directly addresses public safety and storm tracking. The Highlights: 🔶 Technical Complexity: Combined CPLEX with Python’s geospatial plotting. 🔶 Scale: Optimized coverage for the entire state of Texas. 🔶 The "Why": Improved early warning systems for severe weather. Nikhil was introduced to the world of optimization through Professor Michael Ferris's course at UW-Madison. When asked about using GAMSPy, Nikhil said: "What I enjoyed most about working with GAMSPy was its ability to transform complex problems into a solvable and displayable solution... integrating powerful solvers like CPLEX with Python's geospatial data was incredibly useful." Amazing work, Nikhil! We can’t wait to see where your GAMSpy skills take you next! Read the blog 👉 https://lnkd.in/d_6gw3Zh 🔗 Check it out on GitHub: https://lnkd.in/d4E6M6b8 #GAMSPy #Optimization #UWMadison #DataScience #Meteorology #Python #WeatherTech

We’re proud to be part of the UNSEEN project, which applies high-performance computing to large-scale energy energy system modeling. Instead of evaluating a few dozen scenarios, the team analyzed more than 11,000 for the German energy system, enabling a much deeper assessment of uncertainty around costs and security of supply. A meaningful step forward for long-term energy planning. More info below 👇 #HPC #EnergySystems #Optimization #GAMS German Aerospace Center (DLR) Zuse Institute Berlin Technische Universität Berlin Forschungszentrum Jülich University of Graz Gauss Centre for Supercomputing

🥈 2nd Place:  Winner Spotlight: GAMSPy Student Competition: Emilie Lesinski Emilie is in her final year of Industrial Engineering at the University of Wisconsin-Madison. For her project, she used GAMSPy to solve a very real challenge: planning her Post-Graduation Trip across Europe and the UK. The Project: Instead of just guessing a route, Emilie built a multi-objective model to balance travel costs, transit time, and the number of cities visited. She even used "Box Uncertainty" to make the budget robust against 40% spikes in hotel prices. The Highlights: 🔶 Data-Driven: Used real-world pricing from Google Flights and HostelWorld. 🔶 Efficient: Found a 9-city itinerary that fits a $1,700 budget. 🔶 Robust: Designed a plan that stays on track even if travel costs fluctuate. Emilie is graduating this May and heading to McKinsey & Company as a Business Analyst. When asked about using GAMSPy: "What I found most interesting about working with GAMSPy was the ability to model complex optimization problems in an easy to digest way... it gave me the ability to solve a real-world problem that I am passionate about." Congrats Emilie, have a wonderful trip! 👉 Sign up for the GAMS Academic Program - https://lnkd.in/g5eVMSXu #GAMSPy #Optimization #UWMadison #DataScience #TravelTech #IndustrialEngineering

🥉 3rd Place Winner Spotlight: GAMSPy Student Competition 🥉 Big congrats to Koushik M J for taking 3rd place in our student competition! Koushik is an MSc student at RWTH Aachen University and his project, The Energy Negotiator, handles a massive headache for energy transition: balancing neighborhood power grids without invading resident privacy. Using GAMSPy, he built a decentralized framework that coordinates solar panels and batteries through shadow price signals. This smooths out the ‘Duck Curve’ and protects local infrastructure, all while keeping individual household data private and managing financial risk (CVaR). When asked about using GAMSPy, Koushik said: "I really liked the clear and well-structured documentation [of GAMSPy]... it made it very accessible for beginners and easy to start modeling and experimenting quickly." Congrats again on the win, Koushik! 🔗 Check it out on GitHub: https://lnkd.in/dj4rmcTy 👉 Sign up for the GAMS Academic Program - https://lnkd.in/g5eVMSXu #GAMSPy #Optimization #GAMS #SmartGrid #DataScience #RWTHAachen #EnergyTransition