Panua Technologies develops customized high-end software solutions for large-scale prediction, simulation, optimization, and graph analytics.

Our company

Panua - Ipopt

The powerful nonlinear solver, with Panua-Pardiso integrated in it, offering a wide range of performance and robustness improvements.

Download

Panua - Pardiso

The high-performance software for the solution of large sparse symmetric and unsymmetric linear systems of equations.

Download

More about us

Panua Technologies

is a Swiss software company headquartered in Lugano, Switzerland that creates customized high-end software solutions for large-scale prediction, simulation, optimization, and graph analytics. Panua is a spin-off from the Faculty of Informatics at Università della Svizzera italiana . It develops highly flexible software platforms that support various aspects of computational optimization, applied in business-critical problems. Panua bridges the realms of academic algorithmic research and industrial software development to provide accurate and computationally performant graph and data analysis.

Our products

Panua - Ipopt

Ipopt is an especially powerful nonlinear solver, offering a range of state-of-the-art algorithms and options for working with smooth objective and constraint functions in continuous variables.

See more

Panua - Pardiso

The package Pardiso is a thread-safe, high-performance, robust, memory-efficient, and easy-to-use software for solving large sparse linear systems on shared-memory and distributed-memory architectures.

See more

Academic research meets industrial software development

At Panua Technologies, our software solutions are rooted in cutting-edge academic research. As a spin-off from the Institute of Computing (CI) of the Faculty of Informatics at Università della Svizzera italiana, Lugano, Switzerland, we actively collaborate with the academic community on projects related to large-scale optimization, graph analytics, and data analysis. We are committed to empowering the academic world with products that drive scientific progress and spark innovation. Simultaneously, our software solutions have boosted productivity and enhanced efficiency for companies across a wide range of sectors, including aerospace engineering, seismic and reservoir simulations, automotive industries, and more.

Manufacturers and suppliers using Panua software

Newsletter

Optimizing the Pardiso sparse linear solver on Arm architecture

Our Pardiso sparse linear solver is now available on Arm! It's been optimized for Arm NEON and SVE SIMD instructions and performs favorably compared to the Intel MKL version. For more information and comparative results, visit the Arm blogspot announcement.

Learn more

Panua Software used in Grid Optimization Competition Challenge

Panua Technologies is among the sponsors of the Grid Optimization Competition Challenge organized by the Department of Energy (DOE) and both Panua-Pardiso and Panua-Ipopt are available to participants.

Learn more

On cheap entropy-sparsified regression learning

The PNAS paper "On cheap entropy-sparsified regression learning" presents a method for training regression models that is more efficient and effective than previous machine learning approaches. The method, called entropy-sparsified regression learning, uses the algorithm Ipopt in the Panua optimization framework to solve a series of optimization problems. By doing so, it is able to learn complex models with high accuracy while using fewer resources and requiring less data than traditional methods. The paper demonstrates the effectiveness of this method on several real-world datasets, showing that it can be used to improve the performance of machine learning models in a variety of applications. The optimization framework Panua-Ipopt shows promise for increasing efficiency and accuracy in machine learning applications. I. Horenko, E. Vecchi , J. Kardoš, O. Schenk, A. Waechter, T. O’Kane, P. Gagliardini, S. Gerber, On cheap entropy-sparsified regression learning Proceedings of the National Academy of Sciences (PNAS), November 2022, pages 1-12, https://www.pnas.org/doi/10.1073/pnas.2214972120 .

Learn more

Contact