Setting up a z-star vertical coordinate #4087
Unanswered
NoraLoose
asked this question in
Computational science!
Replies: 5 comments 9 replies
-
|
What version of Oceananigans are you using? There was a problem with scalar indexing in the |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
I just registered 0.95.9 which has the new compat for GPUArrays b67ad56 |
Beta Was this translation helpful? Give feedback.
-
|
Oh nice, I'm only seeing this now. |
Beta Was this translation helpful? Give feedback.
-
It is not recommended! I'm surprised we hit |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
looks like it's in the show method of the grid. |
Beta Was this translation helpful? Give feedback.
-
|
ah yes |
Beta Was this translation helpful? Give feedback.
-
|
I'm using I don't seem to be able to upgrade CUDA to |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
I just registered it so it might work now (or in another 30 min) |
Beta Was this translation helpful? Give feedback.
-
|
I have a question separate from the compatibility issue.
|
Beta Was this translation helpful? Give feedback.
2 replies
-
|
I guess the term "faces" suggests absolute heights. But I'm just double-checking. |
Beta Was this translation helpful? Give feedback.
-
|
I believe that's right, ie the input is the same as what you would give the grid constructor. (Also we decided that interface heights made more sense because of the similarity to specifying a uniform grid, ie with an interval, but this can be reconsidered...) |
Beta Was this translation helpful? Give feedback.
-
|
I just tried running this code again: after upgrading to but I still run into the same error: |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
seems that reductions on offset cuda arrays are broken with CUDA 5.6.1 (ClimaOcean) pkg> st
Project ClimaOcean v0.4.4
Status `~/development/ClimaOcean.jl/Project.toml`
[79e6a3ab] Adapt v4.2.0
[179af706] CFTime v0.1.4
[052768ef] CUDA v5.6.1
[6ba0ff68] ClimaSeaIce v0.2.1 `https://github.com/CliMA/ClimaSeaIce.jl.git#ss/new-thermodynamic-timestep`
[9c784101] CubicSplines v0.2.1
[124859b0] DataDeps v0.7.13
[787d08f9] ImageMorphology v0.4.5
[033835bb] JLD2 v0.5.11
[63c18a36] KernelAbstractions v0.9.34
[da04e1cc] MPI v0.20.22
[85f8d34a] NCDatasets v0.14.6
[9e8cae18] Oceananigans v0.95.18
[6fe1bfb0] OffsetArrays v1.15.0
[c2be9673] OrthogonalSphericalShellGrids v0.2.2
[6c6a2e73] Scratch v1.2.1
[d496a93d] SeawaterPolynomials v0.3.5
[90137ffa] StaticArrays v1.9.13
[49b00bb7] SurfaceFluxes v0.12.0
[b60c26fb] Thermodynamics v0.12.9
[ade2ca70] Dates
[f43a241f] Downloads v1.6.0
[de0858da] Printf
[10745b16] Statistics v1.10.0MWE: julia> a = CuArray(zeros(10));
julia> b = OffsetArray(a, -1);
julia> minimum(b)
ERROR: Scalar indexing is disallowed.
Invocation of getindex resulted in scalar indexing of a GPU array.
This is typically caused by calling an iterating implementation of a method.
Such implementations *do not* execute on the GPU, but very slowly on the CPU,
and therefore should be avoided.
If you want to allow scalar iteration, use `allowscalar` or `@allowscalar`
to enable scalar iteration globally or for the operations in question.
Stacktrace:
[1] error(s::String)
@ Base ./error.jl:35
[2] errorscalar(op::String)
@ GPUArraysCore ~/.julia/packages/GPUArraysCore/aNaXo/src/GPUArraysCore.jl:151
[3] _assertscalar(op::String, behavior::GPUArraysCore.ScalarIndexing)
@ GPUArraysCore ~/.julia/packages/GPUArraysCore/aNaXo/src/GPUArraysCore.jl:124
[4] assertscalar(op::String)
@ GPUArraysCore ~/.julia/packages/GPUArraysCore/aNaXo/src/GPUArraysCore.jl:112
[5] getindex
@ ~/.julia/packages/GPUArrays/uiVyU/src/host/indexing.jl:50 [inlined]
[6] getindex
@ ~/.julia/packages/OffsetArrays/HLmxQ/src/OffsetArrays.jl:435 [inlined]
[7] _mapreduce(f::typeof(identity), op::typeof(min), ::IndexLinear, A::OffsetVector{Float64, CuArray{…}})
@ Base ./reduce.jl:438
[8] _mapreduce_dim
@ ./reducedim.jl:367 [inlined]
[9] mapreduce
@ ./reducedim.jl:359 [inlined]
[10] _minimum
@ ./reducedim.jl:1017 [inlined]
[11] _minimum
@ ./reducedim.jl:1016 [inlined]
[12] minimum(a::OffsetVector{Float64, CuArray{Float64, 1, CUDA.DeviceMemory}})
@ Base ./reducedim.jl:1012
[13] top-level scope
@ REPL[57]:1
Some type information was truncated. Use `show(err)` to see complete types. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for narrowing the issue down to an MWE @simone-silvestri! Let's raise an issue upstream on CUDA? |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Oceananigans has a z-star vertical coordinate that conserves tracers (#3956)! 🎉
I just tried to set up an experiment with the new z-star vertical coordinate:
which errors in
I guess I can enable scalar indexing via
but is that recommended?
Beta Was this translation helpful? Give feedback.
All reactions