Toward Exascale Climate Modelling: A Python DSL Approach to ICON&rsquo;s (Icosahedral Non-hydrostatic) Dynamical Core (icon-exclaim v0.2.0)

Dipankar, Anurag; Bianco, Mauro; Bukenberger, Mona; Ehrengruber, Till; Farabullini, Nicoletta; Gopal, Abishek; Hupp, Daniel; Jocksch, Andreas; Kellerhals, Samuel; Kroll, Clarissa A.; Lapillonne, Xavier; Leclair, Matthieu; Luz, Magdalena; Müller, Christoph; Ong, Chia Rui; Osuna, Carlos; Pothapakula, Praveen; Röthlin, Matthias; Sawyer, William; Serafini, Giacomo; Vogt, Hannes; Weber, Ben; Schulthess, Thomas

doi:10.5194/egusphere-2025-4808

Preprints

https://doi.org/10.5194/egusphere-2025-4808

Preprints

14 Oct 2025

| 14 Oct 2025

Status: this preprint is open for discussion and under review for Geoscientific Model Development (GMD).

Toward Exascale Climate Modelling: A Python DSL Approach to ICON’s (Icosahedral Non-hydrostatic) Dynamical Core (icon-exclaim v0.2.0)

Anurag Dipankar, Mauro Bianco, Mona Bukenberger, Till Ehrengruber, Nicoletta Farabullini, Abishek Gopal, Daniel Hupp, Andreas Jocksch, Samuel Kellerhals, Clarissa A. Kroll, Xavier Lapillonne, Matthieu Leclair, Magdalena Luz, Christoph Müller, Chia Rui Ong, Carlos Osuna, Praveen Pothapakula, Matthias Röthlin, William Sawyer, Giacomo Serafini, Hannes Vogt, Ben Weber, and Thomas Schulthess

Abstract. A refactored atmospheric dynamical core of the ICON model implemented in GT4Py, a Python-based domain-specific language designed for performance portability across heterogeneous CPU-GPU architectures, is presented. Integrated within the existing Fortran infrastructure, the GT4Py core achieves throughput slightly exceeding the optimized OpenACC version, reaching up to 213 simulation days per day when using a quarter of CSCS’s ALPS GPUs.

A multi-tiered testing strategy has been implemented to ensure numerical correctness and scientific reliability of the model code. Validation has been performed through global aquaplanet and prescribed sea-surface temperature simulations to demonstrate model’s capability to simulate mesoscale and its interaction with the larger-scale at km-scale grid spacing. This work establishes a foundation for architecture-agnostic ICON global climate and weather model, and highlights poor strong scaling as a potential bottleneck in scaling toward exascale performance.

How to cite. Dipankar, A., Bianco, M., Bukenberger, M., Ehrengruber, T., Farabullini, N., Gopal, A., Hupp, D., Jocksch, A., Kellerhals, S., Kroll, C. A., Lapillonne, X., Leclair, M., Luz, M., Müller, C., Ong, C. R., Osuna, C., Pothapakula, P., Röthlin, M., Sawyer, W., Serafini, G., Vogt, H., Weber, B., and Schulthess, T.: Toward Exascale Climate Modelling: A Python DSL Approach to ICON’s (Icosahedral Non-hydrostatic) Dynamical Core (icon-exclaim v0.2.0), EGUsphere [preprint], https://doi.org/10.5194/egusphere-2025-4808, 2025.

Received: 30 Sep 2025 – Discussion started: 14 Oct 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Status: open (until 09 Dec 2025)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2025-4808', Anonymous Referee #1, 11 Nov 2025 reply
This manuscript presents stage one of a multi-tiered plan to support heterogeneous (mixed CPU/GPU) architectures for running the ICON model. The authors utilize GT4Py, a domain-specific language, to modernize the ICON dynamics core from the existing Fortran code base. The outcome is a more performant code, which is also easier to read and develop compared to the equivalent Fortran implementation. The paper is well written and well reasoned, demonstrating promising results that are on par with the current state of GPU-ready Earth System modeling. I recommend that this manuscript be published, as I have only a few minor questions and technical corrections to suggest.
First, I want to commend the authors for their attention to (a) the hardware-based challenges that arise when running these models at scale, and (b) the importance of robust testing. In my experience, these topics are not typically the most exciting to discuss, but they are essential considerations for any group undertaking a similar effort.

Minor Comments:

Introduction
Paragraph 3: It may be helpful to include node counts when discussing how much of the machine each example used. This additional detail would provide useful context, especially as future machines come online.

Paragraph 6: As noted above, I appreciate the discussion highlighting barriers to running these models at scale.

Section 2
Not strictly necessary, but it may add valuable context for readers if the authors note that Fortran compiler support is increasingly being deprioritized by vendors, which makes supporting legacy codes on new machines more challenging.

Section 3
I may have missed it, but it was unclear whether the plan is to transition entirely away from Fortran after deliverable 3. Could the authors clarify how much of the original Fortran code is expected to remain in the model (e.g., 10%, 25%, or more)?

Section 4
General comment: The authors should verify that each “Listing” is correct and that the code blocks would work as expected.

Section 4.3: If I understood correctly, the ported code was tested to within a tolerance error, and bit-for-bit (BFB) agreement was not strictly enforced. Was any BFB enforcement attempted during porting? If not, could the authors justify their decision not to enforce BFB?

Figure 5: Did the authors conduct experiments with runs well beyond 15 timesteps to confirm that the relative error does indeed stabilize?

Figure 7: Did the authors examine this data using a log-log plot? If so, was the observed trend not quite linear?

Reply
Citation: https://doi.org/10.5194/egusphere-2025-4808-RC1
RC2: 'Comment on egusphere-2025-4808', Anonymous Referee #2, 17 Nov 2025 reply

This is a clear well written paper describing a gt4py implementation

of the ICON dynamical core, running in the existing ICON Fortran

modeling system, enabling k-scale atmospheric simulations on the ALPS

GPU supercomputer. The authors describe their porting approach,

including thorough testing from the kernel level up to full physics

simulations. They provide a sober analysis of the potential of GPUs

and their strong scaling limitations.
I only have minor comments:
1. Section 4.3: what is "the implementation of horizontal blocking"?

Does that refer to the loop blocking in the Fortran loops, (which was

removed in the Python code?)
2. Section 4.3: "...testing is tricky as the results are different due

to rounding..."

The authors have a good port testing strategy in the presence of

roundoff error, but this statement implies that these

rounding differences are unavoidable. The E3SM dycore porting work

(Bertagna et al. GMD 2019 and Bertagna et al. SC2020) showed that it

is possible to obtain BFB agreement between CPUs and GPUs with careful

coding, allowing for a different porting approach which simplifies

some aspects of code porting.
3: Section 5.1:

For the final model, I assume all significant code is running on the

GPUs, with the dycore using gt4py and the physics using openACC. I

believe this is implied, but I didn't see it clearly stated. Were

there any software challenges running the two different GPU

programming models in the same executable?
4. Line 400: "GT4Py synchronization"

I know of two types of synchronization: across MPI nodes, as well as

synchronization among thread teams running on the GPU. Which

is this referring to?
5. Section 5.1

How does the gt4py code compare with the Fortran code on CPUs?

It would be interesting to add CPU-only performance numbers to

Figure 7.

Reply

Citation: https://doi.org/10.5194/egusphere-2025-4808-RC2

Viewed

Total article views: 330 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
250	67	13	330	12	10

HTML: 250
PDF: 67
XML: 13
Total: 330
BibTeX: 12
EndNote: 10

Views and downloads (calculated since 14 Oct 2025)

Month	HTML	PDF	XML	Total
Oct 2025	163	29	8	200
Nov 2025	87	38	5	130

Cumulative views and downloads (calculated since 14 Oct 2025)

Month	HTML	PDF	XML	Total
Oct 2025	163	29	8	200
Nov 2025	87	38	5	130

Viewed (geographical distribution)

Total article views: 323 (including HTML, PDF, and XML) Thereof 323 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 25 Nov 2025

Short summary

Climate models are becoming more detailed and accurate by simulating weather at scales of just a few kilometers. Simulating at km-scale is computationally demanding requiring powerful supercomputers and efficient code. This work presents a refactored dynamical core of a state-of-the-art climate model using a Python-based approach. The refactored code has passed through a sequence of verification and validation demonstrating its usability in performing km-scale global simulations.


Total:	0
HTML:	0
PDF:	0
XML:	0