CMCC Seminar – Bologna
27 February 2025, h. 14:30 CET
The seminar will be held in the CMCC-Bo meeting room (second floor) and via Zoom

Speaker
Simone Silvestri, MIT Department of Earth, Atmospheric, and Planetary Sciences

In this talk, I will present the ocean component of the CliMA’s climate model: Oceananigans, a finite-volume solver of the dynamical equations governing incompressible flows, implemented in the Julia programming language. Oceananigans has been built from the ground up to be fast on GPUs and easy to use. By leveraging Julia, a high-level programming language that combines speed of execution to interpretability, we were able to develop a versatile and modular dynamical core, capable of solving both the nonhydrostatic and hydrostatic incompressible Boussinesq equations. ⁤⁤Moreover, the programmatic user interface of Oceananigans departs from the monolithic, namelist-based interfaces common in traditional ocean models, promoting both usability and extensibility. This makes Oceananigans an ideal platform for developing and testing new parameterizations. Finally, starting from a clean slate allowed us to adopt model design choices customized for GPUs, which often differed from standard ocean modeling practice. Both the model structure and numerical algorithms have been formulated to take advantage of the many parallel cores provided by GPUs, while being mindful that the high-bandwidth memory to which GPUs have access is limited. As a consequence, Oceananigans achieves never-seen performance, capable of simulating a multiple years of global ocean dynamics per wall-clock day at around 10-kilometer horizontal resolution using a handful of A100 GPUs. ⁤⁤This resolution is sufficient to capture the intricate mesoscale eddy field. While ⁤⁤ocean eddies play a key role in the Earth’s climate system, they are typically not simulated by climate models due to their small scale, between 10 and 100 km, which is below the resolution of standard ocean models. By enabling routine eddy-resolving global ocean solutions, Oceananigans holds promise for advancing our understanding of ocean dynamics and improving the predictive capabilities of climate models.