LUMI supercomputer impressed Wärtsilä Finland: superior capacity, cost efficiency, and sustainability

Wärtsilä is a global leader in innovative technologies and lifecycle solutions for the marine and energy markets. Wärtsilä Finland tested LUMI supercomputer for their Computational Fluid Dynamics simulations with superior results in improved efficiency and reduced energy consumption and costs. This article was originally published on the CSC website.

Bulut Tekgül, Wärtsilä’s Combustion & CFD Expert, Thermofluids & Simulations team, works in the company’s R&D and Engineering division. The team utilises Computational Fluid Dynamics (CFD) simulations to understand and improve engine function, propulsion systems, and energy solutions.

– By leveraging advanced CFD techniques, we analyse complex fluid dynamics, combustion processes, and thermal interactions to optimize performance, efficiency, and sustainability. Our work plays a crucial role in driving Wärtsilä’s mission to decarbonize marine and energy industries, ensuring innovative and environmentally friendly solutions for a cleaner future, Tekgül describes.

Complex problems require advanced computing

– We deal with very complex and computationally demanding problems that often include mesh motion, chemical reactions, moving particles, and multi-phase flows. We are constantly aiming for improvements in our cost efficiency and effectiveness in demanding simulations. The scale and complexity of our simulations have to fit within the required timeframes, too, Tekgül says.

To solve some challenges and enhance their computational capabilities, Wärtsilä wanted to explore what the LUMI supercomputer computing infrastructure and CSC expertise has to offer.

– In our collaboration with CSC, we aimed to evaluate how LUMI’s resources could help us perform high-fidelity simulations more efficiently and cost-effectively, Tekgül explains.

Leveraging OpenFOAM

Wärtsilä’s Thermofluids & Simulations team uses OpenFOAM, an open-source software for CFD simulations. OpenFOAM can be used to simulate internal combustion (IC) engines including piston and valve motion – a functionality that Wärtsilä Finland has developed in collaboration with CFD-Direct.

Tekgül notes:

– Our partnership with CFD Direct has been key to unlocking the full potential of OpenFOAM. Combined with LUMI’s computational power, we can tackle large-scale simulations effectively.

LUMI supercomputer for business use

It is not a given that one can incorporate supercomputing into their computing repertoire. Tekgül had the advantage of having previously used CSC’s computing services: initially Taito and Sisu supercomputers, and later Puhti and Mahti for reactive Computational Fluid Dynamics problems during his doctoral studies at Aalto University.

He was well aware that LUMI’s unparalleled computing power and advanced simulation capabilities had earned recognition not only from his research colleagues but also from the media. Once he became aware that CSC was also offering solutions for industrial use, Wärtsilä decided to test LUMI via CSC’s Try and Buy model.

– The uptake process was straightforward. After the creation of the user accounts and transfer of the billing units, we were able to compile the custom libraries. We still needed some help from CSC experts, and after their quick response, we were able to start our test runs, Tekgül says.

Cost efficiency boost with reliable support

– We were really positively surprised by the results. The improved efficiency was clearly visible in our test runs. We were able to run our simulations much faster, meeting tight project deadlines and accelerating our R&D processes, Tekgül says.

– We also assessed the cost efficiency of our in-house infrastructure versus CSC’s services. We compared the costs, scalability, and performance benefits of each approach, which provided us with a clear understanding of the potential advantages of using high-performance computing. CSC’s responsive service desk ensured that any technical challenges were addressed promptly, minimizing downtime and enhancing overall productivity.

The test runs also serve Wärtsilä’s ambitions of becoming carbon neutral by 2030. To achieve this goal, the company invests in developing new technologies, large-scale simulations and the increased use of machine learning. All this requires more computing power.

Maximizing simulations, minimizing energy consumption and costs

– Our focus is on ramping up our industrial R&D computations in LUMI with the applications already identified. Obviously, our goal is to maximize the use of simulations and minimize carbon-intensive energy consumption and costs. If we are successful, we should be able to leverage the use of any capacity offered to us, Tekgül says.

– We also need to adopt new technology more quickly, starting with GPU computing, continuing with Artificial Intelligence models and hopefully soon, using quantum computing, he concludes.

Read also:

Did you know that up to 20% of LUMI’s resources are reserved for industrial users? Learn more in the Get started section.

Author: EuroCC Finland