The State of InfiniteOpt.jl | Joshua Pulsipher | JuMP-dev 2025 @TheJuliaLanguage

Uploaded November 2025 | Updated September 2026, 2 weeks ago

InfiniteOpt.jl provides an algebraic modeling language for intuitively model and rapidly solve infinite-dimensional optimization problems which arise in numerous application areas such as optimal control, two-stage stochastic programming, and PDE-constrained optimization. This talk will detail key advancements recently made to InfiniteOpt that include specialized support for infinite-dimensional generalized disjunctive programming problems, tailored GPU-workflows for rapid nonlinear model predictive control, and embedding infinite-dimensional surrogate models. A roadmap for and perspective of future development will also be shared.

The State of InfiniteOpt.jl | Joshua Pulsipher | JuMP-dev 2025

QUBO.jl | Maciel Xavier | JuliaCon Global 2025

Advanced Model Composition in ApplicationDrivenLearning.jl | Giovanni Amorim

Computational Quantum Chemistry with Sparse Matrix Algorithms | Madureira | JuliaCon Global 2025

Computer-Aided Design of Quantum Integrated Circuits | Peairs | JuliaCon 2024

UniversalDiffEq.jl: applying SciML to ecology | H Buckner | JuliaCon Global 2025

FerriteCon 2025 Basavesh Yaraguntappa: Variational Minimization Formulation

Muschel N1 | JuliaCon Global 2026 | Day 2

TrixiCUDA.jl: CUDA Support for Solving Hyperbolic PDEs on GPU | Xie | JuliaCon Global 2025

GraphsOfConvexSets.jl: Optimisation over Graphs of Convex Sets with JuMP | Charles van Hees

$Accelerating Fermi Operator Expansion: ML-Inspired Methods | Zhang | JuliaCon Global 2025 Accelerating Fermi Operator Expansion: ML-Inspired Methods by Qi Zhang PreTalx: https://pretalx.com/juliacon-2025/talk/88WEPL/ Calculating electronic properties of materials using the density matrix formalism enables linear scaling with system size, compared to the cubic scaling of direct diagonalization. To obtain density matrices, one typically applies the Fermi–Dirac operator to the Hamiltonian of a given system. The second-order spectral projection (SP2) scheme has proven effective across a broad range of temperatures. We generalize this SP2 method, drawing inspiration from machine learning models, to recursively expand the Fermi–Dirac operator. This approach accurately captures properties such as fractional occupation numbers and electronic entropy while significantly reducing computational cost by replacing highly nonlinear matrix functions with iterative matrix multiplications. Performance further improves by running these calculations on GPUs. In this presentation, we show how Julia enables rapid prototyping of this algorithm and the subsequent design of a user-friendly API. We share strategies for optimizing performance (e.g., removing type instabilities, reducing allocations, and avoiding code duplication) with the help of robust benchmarking tools from the Julia community. We have been pleasantly surprised to see that writing GPU-compatible Julia code adds minimal overhead compared to calling CUDA C++ directly, while substantially improving readability—particularly when leveraging tensor cores. We also take advantage of automatic differentiation packages to generalize our approach to differentiating physically meaningful functions. Test cases for this model use a rich ecosystem of random distribution generation code, showcasing the flexibility and power of Julia for computational chemistry applications.$