The paper advances the science by introducing two new metrics, Peak Ratio Crossover (PRC) and Peek-to-Peek Ratio (PPR), which allow for a better comparison between CPU and GPU performance of a given application. The authors demonstrate the benefits of using these metrics on a single-column parametrisation of turbulence and clouds, Cloud Layers Unified by Binormals  (CLUBB), which exposes several level of parallelism that can be exploited differently by heterogenous architectures. Several use cases show the impact of batch size, precision, asynchronous execution, device type and coding optimisations on the application's throughput, expressed in columns per second. The paper is well written and the claims are well supported by experiments and profiling data which naturally drive the conclusions. I would recommend the publication of the manuscript. 

There are a few comments, listed below, which could be tackled by the authors in a minor revision:

• The GPU in the title is a broad topic while most of the findings are based on NVIDIA GPUs experiments. Would be interesting to see if all the conclusions still hold on MI250x (given different warp size and different generated assembly code). That is more for the future, not needed to be added for the current manuscript.
• Line 107: "This setup .. closely mirrors the execution model.." - This is usually only partially true because a standalone has a smaller memory footprint in terms of the instruction cache. 
• Fig 3 / page 9 could be improved if the plots with the same number of levels would have the same colour, since the vertical levels are in focus here. E.g. AMD7763_2x128_34nz and A100_4x4_34nz.
• Section 5.4 OpenACC vs OpenMP - Firstly, the experiments were done with the NVIDIA compiler which is known to favour OpenACC (e.g. the amount of optimisations for it is significantly larger than for OpenMP), so double-checking the results with other compiler could bring new information. Secondly, Fortran + OpenMP is known to perform better than Fortran + OpenACC on AMD hardware (using AMD ROCm compiler stack),  so without a similar experiment on AMD, I would rephrase the findings in lines 297-306 as limited to NVIDIA hardware using NVIDIA compiler family. 
• Since the manuscript is not anonymised, maybe the authors can write their full names in lines 518-520.

This paper gives a detailed performance analysis of the CLUBB cloud

and turbulence parameterization, including a new GPU openACC based

port.  CLUBB is an important and expensive component used by modern

atmospheric models.  The performance work described in this paper is

excellent.  The authors give a thorough evaluation of both GPU and CPU

performance as a function of workload ( number of vertical physics

columns per device), compilers, precision, vertical levels and loop

structure.  They go to great lengths to prevent a fair CPU vs GPU

comparison, comparing well performing CPU code to similarly well

performing GPU code.  They use several metrics (PPR, PRC, ATR, DRC) and

carefully discuss the strengths and weaknesses of CPUs and GPUs.  The

thoroughness and fairness of the paper puts it above many other papers

which present misleading GPU speedup numbers.

The paper is well written with clear arguments.

I only have two minor suggestions:

1. Section 2: line 92:

The wording of the sentence containing the Sun et al 2023 reference

could be clarified.  At first reading, I assumed the authors were saying

that the column loop changes in CLUBB were described in Sun et al 2023,

but I think the CLUBB changes are due to this work and Sun et al 2023

was discussing similar work in a different parameterization (PUMAS).

2. Conclusions:

I think most readers will be interested in how the CLUBB performance

numbers will impact global atmospheric models, in typical regimes with

are running at the limit of strong scaling to get the maximum possible

throughput, or running on fewer nodes to maximize efficiency and

maximize ensemble throughput.  The author's detailed benchmarks and

metrics will allow any motivated reader to answer this question.  But

I think it would save the readers a little time if the authors added a

short discussion explaining the relation between PPR and PRC and these

strong scaling or maximum efficiency global model configurations.