What is your role in ESiWACE?
Co-leading WP2. WP2 aims at evaluating and establishing new technologies. There are a number of technologies that emerged already in the past decade and are being actively developed that are highly interesting for our community. The weather and climate community has a large code base of production applications. Therefore, the adoption of new technologies is a long process, and often iterative, i.e. the adoption of certain technologies in our models influences the further development of the technology.
ESiWACE will play a key role in the use of, among others, domain specific languages for weather and climate applications, containers for (low latency) MPI distributed HPC applications or solutions to concurrent runs of model components.
What do you appreciate most about your work?
I enjoy the most the international collaboration and working together with a motivated and highly talented colleagues, as well as the goals of ESiWACE of giving a service to the community. While model developers can focus on science, we focus our work in WP2 to pave the way to efficient use of complex and hybrid machines and find solutions for long-term sustainable models.
Which question in climate and weather research interests you the most?
I am most interested in the technology and how we can adopt new technologies that will allow our community to keep using efficiently the next generation of hybrid supercomputer systems. In the past years we have been seeing incredible progress in hardware and an explosion of computing architectures. This is at the same time a great opportunity for our community to make use of the latest and most powerful processor and accelerators for scientific simulations. At the same time it poses a great challenge for the adoption and use of new computing architectures, while retaining a readable and long-term sustainable software code base.
What drives you and what do you want to achieve?
I would like to see some of the technologies that we are evaluating today becoming, in the future, common programming paradigms in weather and climate model developments, allowing us to rapidly adapt to the continuously changing environment in hardware driven by the market. So that we can focus on what really matters, which is science productivity.
Carlos Osuna
Co-lead of the computing for numerical development team
MeteoSwiss
Zurich, Switzerland
e-mail: Carlos.Osuna@meteoswiss.ch