use Float32 as default (#70)

* add support for Metal.jl

* use Float32 as default

* revert changes for Metal.jl

* bug fix

Author: zhenwu0728

Manifest.toml

@@ -1,8 +1,8 @@
 # This file is machine-generated - editing it directly is not advised
 
-julia_version = "1.10.2"
+julia_version = "1.10.3"
 manifest_format = "2.0"
-project_hash = "355ce7a1dcdff945cffe1b1dd3a7f5c1a045276b"
+project_hash = "1078913c9829ff096c73958654eaca165e82cef9"
 
 [[deps.AbstractFFTs]]
 deps = ["LinearAlgebra"]
@@ -43,9 +43,9 @@ version = "0.1.0"
 
 [[deps.BFloat16s]]
 deps = ["LinearAlgebra", "Printf", "Random", "Test"]
-git-tree-sha1 = "dbf84058d0a8cbbadee18d25cf606934b22d7c66"
+git-tree-sha1 = "2c7cc21e8678eff479978a0a2ef5ce2f51b63dff"
 uuid = "ab4f0b2a-ad5b-11e8-123f-65d77653426b"
-version = "0.4.2"
+version = "0.5.0"
 
 [[deps.Base64]]
 uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
@@ -57,9 +57,9 @@ version = "0.5.0"
 
 [[deps.CUDA]]
 deps = ["AbstractFFTs", "Adapt", "BFloat16s", "CEnum", "CUDA_Driver_jll", "CUDA_Runtime_Discovery", "CUDA_Runtime_jll", "Crayons", "DataFrames", "ExprTools", "GPUArrays", "GPUCompiler", "KernelAbstractions", "LLVM", "LLVMLoopInfo", "LazyArtifacts", "Libdl", "LinearAlgebra", "Logging", "NVTX", "Preferences", "PrettyTables", "Printf", "Random", "Random123", "RandomNumbers", "Reexport", "Requires", "SparseArrays", "StaticArrays", "Statistics"]
-git-tree-sha1 = "baa8ea7a1ea63316fa3feb454635215773c9c845"
+git-tree-sha1 = "4e33522a036b39fc6f5cb7447ae3b28eb8fbe99b"
 uuid = "052768ef-5323-5732-b1bb-66c8b64840ba"
-version = "5.2.0"
+version = "5.3.3"
 
     [deps.CUDA.extensions]
     ChainRulesCoreExt = "ChainRulesCore"
@@ -71,27 +71,27 @@ version = "5.2.0"
 
 [[deps.CUDA_Driver_jll]]
 deps = ["Artifacts", "JLLWrappers", "LazyArtifacts", "Libdl", "Pkg"]
-git-tree-sha1 = "d01bfc999768f0a31ed36f5d22a76161fc63079c"
+git-tree-sha1 = "dc172b558adbf17952001e15cf0d6364e6d78c2f"
 uuid = "4ee394cb-3365-5eb0-8335-949819d2adfc"
-version = "0.7.0+1"
+version = "0.8.1+0"
 
 [[deps.CUDA_Runtime_Discovery]]
 deps = ["Libdl"]
-git-tree-sha1 = "2cb12f6b2209f40a4b8967697689a47c50485490"
+git-tree-sha1 = "38f830504358e9972d2a0c3e5d51cb865e0733df"
 uuid = "1af6417a-86b4-443c-805f-a4643ffb695f"
-version = "0.2.3"
+version = "0.2.4"
 
 [[deps.CUDA_Runtime_jll]]
 deps = ["Artifacts", "CUDA_Driver_jll", "JLLWrappers", "LazyArtifacts", "Libdl", "TOML"]
-git-tree-sha1 = "8e25c009d2bf16c2c31a70a6e9e8939f7325cc84"
+git-tree-sha1 = "4ca7d6d92075906c2ce871ea8bba971fff20d00c"
 uuid = "76a88914-d11a-5bdc-97e0-2f5a05c973a2"
-version = "0.11.1+0"
+version = "0.12.1+0"
 
 [[deps.ColorTypes]]
 deps = ["FixedPointNumbers", "Random"]
-git-tree-sha1 = "eb7f0f8307f71fac7c606984ea5fb2817275d6e4"
+git-tree-sha1 = "b10d0b65641d57b8b4d5e234446582de5047050d"
 uuid = "3da002f7-5984-5a60-b8a6-cbb66c0b333f"
-version = "0.11.4"
+version = "0.11.5"
 
 [[deps.Colors]]
 deps = ["ColorTypes", "FixedPointNumbers", "Reexport"]
@@ -101,9 +101,9 @@ version = "0.12.10"
 
 [[deps.Compat]]
 deps = ["TOML", "UUIDs"]
-git-tree-sha1 = "c955881e3c981181362ae4088b35995446298b80"
+git-tree-sha1 = "b1c55339b7c6c350ee89f2c1604299660525b248"
 uuid = "34da2185-b29b-5c13-b0c7-acf172513d20"
-version = "4.14.0"
+version = "4.15.0"
 weakdeps = ["Dates", "LinearAlgebra"]
 
     [deps.Compat.extensions]
@@ -112,7 +112,7 @@ weakdeps = ["Dates", "LinearAlgebra"]
 [[deps.CompilerSupportLibraries_jll]]
 deps = ["Artifacts", "Libdl"]
 uuid = "e66e0078-7015-5450-92f7-15fbd957f2ae"
-version = "1.1.0+0"
+version = "1.1.1+0"
 
 [[deps.ConstructionBase]]
 deps = ["LinearAlgebra"]
@@ -146,9 +146,9 @@ version = "1.6.1"
 
 [[deps.DataStructures]]
 deps = ["Compat", "InteractiveUtils", "OrderedCollections"]
-git-tree-sha1 = "0f4b5d62a88d8f59003e43c25a8a90de9eb76317"
+git-tree-sha1 = "1d0a14036acb104d9e89698bd408f63ab58cdc82"
 uuid = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
-version = "0.18.18"
+version = "0.18.20"
 
 [[deps.DataValueInterfaces]]
 git-tree-sha1 = "bfc1187b79289637fa0ef6d4436ebdfe6905cbd6"
@@ -190,9 +190,9 @@ uuid = "9fa8497b-333b-5362-9e8d-4d0656e87820"
 
 [[deps.GPUArrays]]
 deps = ["Adapt", "GPUArraysCore", "LLVM", "LinearAlgebra", "Printf", "Random", "Reexport", "Serialization", "Statistics"]
-git-tree-sha1 = "47e4686ec18a9620850bad110b79966132f14283"
+git-tree-sha1 = "68e8ff56a4a355a85d2784b94614491f8c900cde"
 uuid = "0c68f7d7-f131-5f86-a1c3-88cf8149b2d7"
-version = "10.0.2"
+version = "10.1.0"
 
 [[deps.GPUArraysCore]]
 deps = ["Adapt"]
@@ -202,9 +202,9 @@ version = "0.1.6"
 
 [[deps.GPUCompiler]]
 deps = ["ExprTools", "InteractiveUtils", "LLVM", "Libdl", "Logging", "Scratch", "TimerOutputs", "UUIDs"]
-git-tree-sha1 = "a846f297ce9d09ccba02ead0cae70690e072a119"
+git-tree-sha1 = "1600477fba37c9fc067b9be21f5e8101f24a8865"
 uuid = "61eb1bfa-7361-4325-ad38-22787b887f55"
-version = "0.25.0"
+version = "0.26.4"
 
 [[deps.InlineStrings]]
 deps = ["Parsers"]
@@ -258,9 +258,9 @@ version = "0.9.18"
 
 [[deps.LLVM]]
 deps = ["CEnum", "LLVMExtra_jll", "Libdl", "Preferences", "Printf", "Requires", "Unicode"]
-git-tree-sha1 = "ab01dde107f21aa76144d0771dccc08f152ccac7"
+git-tree-sha1 = "839c82932db86740ae729779e610f07a1640be9a"
 uuid = "929cbde3-209d-540e-8aea-75f648917ca0"
-version = "6.6.2"
+version = "6.6.3"
 weakdeps = ["BFloat16s"]
 
     [deps.LLVM.extensions]
@@ -337,9 +337,9 @@ version = "2.28.2+1"
 
 [[deps.Missings]]
 deps = ["DataAPI"]
-git-tree-sha1 = "f66bdc5de519e8f8ae43bdc598782d35a25b1272"
+git-tree-sha1 = "ec4f7fbeab05d7747bdf98eb74d130a2a2ed298d"
 uuid = "e1d29d7a-bbdc-5cf2-9ac0-f12de2c33e28"
-version = "1.1.0"
+version = "1.2.0"
 
 [[deps.Mmap]]
 uuid = "a63ad114-7e13-5084-954f-fe012c677804"
@@ -558,9 +558,9 @@ uuid = "a759f4b9-e2f1-59dc-863e-4aeb61b1ea8f"
 version = "0.5.23"
 
 [[deps.TranscodingStreams]]
-git-tree-sha1 = "71509f04d045ec714c4748c785a59045c3736349"
+git-tree-sha1 = "5d54d076465da49d6746c647022f3b3674e64156"
 uuid = "3bb67fe8-82b1-5028-8e26-92a6c54297fa"
-version = "0.10.7"
+version = "0.10.8"
 weakdeps = ["Random", "Test"]
 
     [deps.TranscodingStreams.extensions]

Project.toml

@@ -6,6 +6,7 @@ version = "0.7.0"
 [deps]
 Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
 CUDA = "052768ef-5323-5732-b1bb-66c8b64840ba"
+GPUArrays = "0c68f7d7-f131-5f86-a1c3-88cf8149b2d7"
 JLD2 = "033835bb-8acc-5ee8-8aae-3f567f8a3819"
 KernelAbstractions = "63c18a36-062a-441e-b654-da1e3ab1ce7c"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
@@ -30,6 +31,5 @@ Coverage = "a2441757-f6aa-5fb2-8edb-039e3f45d037"
 MeshArrays = "cb8c808f-1acf-59a3-9d2b-6e38d009f683"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
-
 [targets]
 test = ["BenchmarkTools", "Coverage", "MeshArrays", "Test"]

src/Architectures.jl

@@ -5,6 +5,7 @@ export array_type, rng_type
 export device
 
 using CUDA
+using GPUArrays
 using KernelAbstractions
 using CUDA.CUDAKernels
 using Random
@@ -23,10 +24,11 @@ struct CPU <: Architecture end
 
 """
     GPU <: Architecture
-Run PlanktonIndividuals on one GPU node.
+Run PlanktonIndividuals on one CUDA GPU node.
 """
 struct GPU <: Architecture end
 
+
 device(::CPU) = KernelAbstractions.CPU()
 device(::GPU) = CUDABackend()
 
@@ -36,4 +38,4 @@ array_type(::GPU) = CuArray
 rng_type(::CPU) = MersenneTwister()
 rng_type(::GPU) = CURAND.default_rng()
 
-end
+end

src/Biogeochemistry/Advection/DST3FL.jl

@@ -1,14 +1,14 @@
 ##### halo points ≥ 2 NEEDED!!
-const θmax = 1.0e20
+const θmax = 1.0f20
 
 ##### calculate CFL number: c=uΔt/Lx
 @inline CFLx(i, j, k, g::AbstractGrid, u, ΔT) = @inbounds abs(u[i, j, k] * ΔT / ΔxC(i, j, k, g))
 @inline CFLy(i, j, k, g::AbstractGrid, v, ΔT) = @inbounds abs(v[i, j, k] * ΔT / ΔyC(i, j, k, g))
 @inline CFLz(i, j, k, g::AbstractGrid, w, ΔT) = @inbounds abs(w[i, j, k] * ΔT / ΔzC(i, j, k, g))
 
 ##### calculate d₀ and d₁
-@inline d0(CFL) = @inbounds (2.0 - CFL) * (1.0 - CFL) / 6.0
-@inline d1(CFL) = @inbounds (1.0 - CFL * CFL) / 6.0
+@inline d0(CFL) = @inbounds (2.0f0 - CFL) * (1.0f0 - CFL) / 6.0f0
+@inline d1(CFL) = @inbounds (1.0f0 - CFL * CFL) / 6.0f0
 
 ##### calculate volume transport, unit: m³/s
 @inline Trans_x(i, j, k, g::AbstractGrid, u) = @inbounds Ax(i, j, k, g) * u[i, j, k]
@@ -36,38 +36,38 @@ const θmax = 1.0e20
 
 ##### calculate Ψ⁺ and Ψ⁻
 @inline Ψx⁺(i, j, k, g::AbstractGrid, u, c, ΔT) = 
-            max(0.0, min(min(1.0, d0(CFLx(i, j, k, g, u, ΔT)) + d1(CFLx(i, j, k, g, u, ΔT)) * θx⁺(i, j, k, c, g)), 
-                         θx⁺(i, j, k, c, g) * (1.0 - CFLx(i, j, k, g, u, ΔT)) / (CFLx(i, j, k, g, u, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLx(i, j, k, g, u, ΔT)) + d1(CFLx(i, j, k, g, u, ΔT)) * θx⁺(i, j, k, c, g)), 
+                         θx⁺(i, j, k, c, g) * (1.0f0 - CFLx(i, j, k, g, u, ΔT)) / (CFLx(i, j, k, g, u, ΔT) + 1.0f-20)))
 
 @inline Ψx⁻(i, j, k, g::AbstractGrid, u, c, ΔT) = 
-            max(0.0, min(min(1.0, d0(CFLx(i, j, k, g, u, ΔT)) + d1(CFLx(i, j, k, g, u, ΔT)) * θx⁻(i, j, k, c, g)), 
-                         θx⁻(i, j, k, c, g) * (1.0 - CFLx(i, j, k, g, u, ΔT)) / (CFLx(i, j, k, g, u, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLx(i, j, k, g, u, ΔT)) + d1(CFLx(i, j, k, g, u, ΔT)) * θx⁻(i, j, k, c, g)), 
+                         θx⁻(i, j, k, c, g) * (1.0f0 - CFLx(i, j, k, g, u, ΔT)) / (CFLx(i, j, k, g, u, ΔT) + 1.0f-20)))
 
 @inline Ψy⁺(i, j, k, g::AbstractGrid, v, c, ΔT) = 
-            max(0.0, min(min(1.0, d0(CFLy(i, j, k, g, v, ΔT)) + d1(CFLy(i, j, k, g, v, ΔT)) * θy⁺(i, j, k, c, g)), 
-                         θy⁺(i, j, k, c, g) * (1.0 - CFLy(i, j, k, g, v, ΔT)) / (CFLy(i, j, k, g, v, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLy(i, j, k, g, v, ΔT)) + d1(CFLy(i, j, k, g, v, ΔT)) * θy⁺(i, j, k, c, g)), 
+                         θy⁺(i, j, k, c, g) * (1.0f0 - CFLy(i, j, k, g, v, ΔT)) / (CFLy(i, j, k, g, v, ΔT) + 1.0f-20)))
 
 @inline Ψy⁻(i, j, k, g::AbstractGrid, v, c, ΔT) =
-            max(0.0, min(min(1.0, d0(CFLy(i, j, k, g, v, ΔT)) + d1(CFLy(i, j, k, g, v, ΔT)) * θy⁻(i, j, k, c, g)), 
-                         θy⁻(i, j, k, c, g) * (1.0 - CFLy(i, j, k, g, v, ΔT)) / (CFLy(i, j, k, g, v, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLy(i, j, k, g, v, ΔT)) + d1(CFLy(i, j, k, g, v, ΔT)) * θy⁻(i, j, k, c, g)), 
+                         θy⁻(i, j, k, c, g) * (1.0f0 - CFLy(i, j, k, g, v, ΔT)) / (CFLy(i, j, k, g, v, ΔT) + 1.0f-20)))
 
 @inline Ψz⁺(i, j, k, g::AbstractGrid, w, c, ΔT) = 
-            max(0.0, min(min(1.0, d0(CFLz(i, j, k, g, w, ΔT)) + d1(CFLz(i, j, k, g, w, ΔT)) * θz⁺(i, j, k, c, g)), 
-                         θz⁺(i, j, k, c, g) * (1.0 - CFLz(i, j, k, g, w, ΔT)) / (CFLz(i, j, k, g, w, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLz(i, j, k, g, w, ΔT)) + d1(CFLz(i, j, k, g, w, ΔT)) * θz⁺(i, j, k, c, g)), 
+                         θz⁺(i, j, k, c, g) * (1.0f0 - CFLz(i, j, k, g, w, ΔT)) / (CFLz(i, j, k, g, w, ΔT) + 1.0f-20)))
 
 @inline Ψz⁻(i, j, k, g::AbstractGrid, w, c, ΔT) = 
-            max(0.0, min(min(1.0, d0(CFLz(i, j, k, g, w, ΔT)) + d1(CFLz(i, j, k, g, w, ΔT)) * θz⁻(i, j, k, c, g)), 
-                         θz⁻(i, j, k, c, g) * (1.0 - CFLz(i, j, k, g, w, ΔT)) / (CFLz(i, j, k, g, w, ΔT) + 1.0e-20)))
+            max(0.0f0, min(min(1.0f0, d0(CFLz(i, j, k, g, w, ΔT)) + d1(CFLz(i, j, k, g, w, ΔT)) * θz⁻(i, j, k, c, g)), 
+                         θz⁻(i, j, k, c, g) * (1.0f0 - CFLz(i, j, k, g, w, ΔT)) / (CFLz(i, j, k, g, w, ΔT) + 1.0f-20)))
 
 ##### advection flux
 @inline adv_flux_x(i, j, k, g::AbstractGrid, u, c, ΔT) = 
-    0.5 * (Trans_x(i, j, k, g, u) + abs(Trans_x(i, j, k, g, u))) * (c[i-1, j, k] + Ψx⁺(i, j, k, g, u, c, ΔT) * δx⁰(i, j, k, c, g)) +
-    0.5 * (Trans_x(i, j, k, g, u) - abs(Trans_x(i, j, k, g, u))) * (c[i,   j, k] - Ψx⁻(i, j, k, g, u, c, ΔT) * δx⁰(i, j, k, c, g))
+    0.5f0 * (Trans_x(i, j, k, g, u) + abs(Trans_x(i, j, k, g, u))) * (c[i-1, j, k] + Ψx⁺(i, j, k, g, u, c, ΔT) * δx⁰(i, j, k, c, g)) +
+    0.5f0 * (Trans_x(i, j, k, g, u) - abs(Trans_x(i, j, k, g, u))) * (c[i,   j, k] - Ψx⁻(i, j, k, g, u, c, ΔT) * δx⁰(i, j, k, c, g))
 
 @inline adv_flux_y(i, j, k, g::AbstractGrid, v, c, ΔT) =
-    0.5 * (Trans_y(i, j, k, g, v) + abs(Trans_y(i, j, k, g, v))) * (c[i, j-1, k] + Ψy⁺(i, j, k, g, v, c, ΔT) * δy⁰(i, j, k, c, g)) +
-    0.5 * (Trans_y(i, j, k, g, v) - abs(Trans_y(i, j, k, g, v))) * (c[i, j,   k] - Ψy⁻(i, j, k, g, v, c, ΔT) * δy⁰(i, j, k, c, g))
+    0.5f0 * (Trans_y(i, j, k, g, v) + abs(Trans_y(i, j, k, g, v))) * (c[i, j-1, k] + Ψy⁺(i, j, k, g, v, c, ΔT) * δy⁰(i, j, k, c, g)) +
+    0.5f0 * (Trans_y(i, j, k, g, v) - abs(Trans_y(i, j, k, g, v))) * (c[i, j,   k] - Ψy⁻(i, j, k, g, v, c, ΔT) * δy⁰(i, j, k, c, g))
 
 @inline adv_flux_z(i, j, k, g::AbstractGrid, w, c, ΔT) =
-    0.5 * (Trans_z(i, j, k, g, w) + abs(Trans_z(i, j, k, g, w))) * (c[i, j, k  ] + Ψz⁻(i, j, k, g, w, c, ΔT) * δz⁰(i, j, k, c, g)) +
-    0.5 * (Trans_z(i, j, k, g, w) - abs(Trans_z(i, j, k, g, w))) * (c[i, j, k-1] - Ψz⁺(i, j, k, g, w, c, ΔT) * δz⁰(i, j, k, c, g))
+    0.5f0 * (Trans_z(i, j, k, g, w) + abs(Trans_z(i, j, k, g, w))) * (c[i, j, k  ] + Ψz⁻(i, j, k, g, w, c, ΔT) * δz⁰(i, j, k, c, g)) +
+    0.5f0 * (Trans_z(i, j, k, g, w) - abs(Trans_z(i, j, k, g, w))) * (c[i, j, k-1] - Ψz⁺(i, j, k, g, w, c, ΔT) * δz⁰(i, j, k, c, g))

src/Biogeochemistry/Advection/tracer_diffusion.jl

@@ -5,9 +5,9 @@
 
 ##### Laplacian diffusion operator
 @inline function κ∇²(i, j, k, grid, κˣ, κʸ, κᶻ, c)
-    return 1/volume(i, j, k, grid) * (δx⁺(i, j, k, grid, diffusive_flux_x, κˣ, c) +
-                                      δy⁺(i, j, k, grid, diffusive_flux_y, κʸ, c) +
-                                      δz⁺(i, j, k, grid, diffusive_flux_z, κᶻ, c))
+    return 1.0f0/volume(i, j, k, grid) * (δx⁺(i, j, k, grid, diffusive_flux_x, κˣ, c) +
+                                          δy⁺(i, j, k, grid, diffusive_flux_y, κʸ, c) +
+                                          δz⁺(i, j, k, grid, diffusive_flux_z, κᶻ, c))
 end
 
 ##### calculate the tendency by diffusion for tracer c

src/Biogeochemistry/nutrient_fields.jl

@@ -53,7 +53,7 @@ function generate_nutrients(arch::Architecture, g::AbstractGrid,
                 end
 
                 if size(tmp) == (g.Nx, g.Ny, g.Nz)
-                    @views @. nut[name].data[g.Hx+1:g.Hx+g.Nx, g.Hy+1:g.Hy+g.Ny, g.Hz+1:g.Hz+g.Nz] = tmp[:,:,:]
+                    @views @. nut[name].data[g.Hx+1:g.Hx+g.Nx, g.Hy+1:g.Hy+g.Ny, g.Hz+1:g.Hz+g.Nz] = FT.(tmp[:,:,:])
                 else
                     throw(ArgumentError("NUT_INIT:  grid mismatch"))
                 end
@@ -63,7 +63,7 @@ function generate_nutrients(arch::Architecture, g::AbstractGrid,
                 end
                 lower = FT(1.0 - source.rand_noise[name])
                 upper = FT(1.0 + source.rand_noise[name])
-                nut[name].data .= fill(FT(source.initial_condition[name]),total_size) .* rand(lower:1e-4:upper, total_size) |> array_type(arch)
+                nut[name].data .= fill(FT(source.initial_condition[name]),total_size) .* rand(lower:1.0f-4:upper, total_size) |> array_type(arch)
             end
         end
 

src/Biogeochemistry/nutrient_forcings.jl

@@ -1,7 +1,7 @@
 ##### calculate simple remineralization of DOM and POM as well as a simple nitrification
 function nut_forcing!(F, nut_temp, nut, params, ΔT)
     for name in nut_names
-        @inbounds nut_temp[name].data .= max.(0.0, nut[name].data)
+        @inbounds nut_temp[name].data .= max.(0.0f0, nut[name].data)
     end
 
     @inbounds F.DIC.data .= F.DIC.data .+ nut_temp.DOC.data .* params["kDOC"] .* ΔT

src/Biogeochemistry/nutrient_update.jl

@@ -37,17 +37,4 @@ function nut_update!(nutrients, Gcs, nut_temp, arch::Architecture, g::AbstractGr
     apply_tendency!(nutrients, Gcs, consume, ΔT, g, arch)
 
     fill_halo_nut!(nutrients, g)
-end
-
-# function nut_update!(nutrients, Gcs, nut_temp, arch::Architecture, g::AbstractGrid, params, consume, ΔT, iter)
-#     # compute biogeochemical forcings of nutrients,for each time step
-#     nut_forcing!(Gcs, nut_temp, nutrients, params, ΔT)
-
-#     ##### apply boundary conditions
-#     apply_bcs!(Gcs, nutrients, g, iter, ΔT, arch)
-
-#     # apply forcing tendency
-#     apply_tendency!(nutrients, Gcs, consume, ΔT, g, arch)
-
-#     fill_halo_nut!(nutrients, g)
-# end
+end

src/Diagnostics/Diagnostics.jl

@@ -7,7 +7,7 @@ export tracer_avail_diags, plank_avail_diags
 using CUDA
 using KernelAbstractions
 
-using PlanktonIndividuals.Architectures: device, Architecture, GPU, CPU, array_type
+using PlanktonIndividuals.Architectures: device, Architecture, array_type
 using PlanktonIndividuals.Grids
 using PlanktonIndividuals.Fields
 

src/Fields/Fields.jl

@@ -39,7 +39,7 @@ end
 
 function zero_fields!(a)
     for i in 1:length(a)
-        @inbounds a[i].data .= 0.0
+        @inbounds a[i].data .= 0.0f0
     end
 end
 

src/Fields/boundary_conditions.jl

@@ -32,9 +32,9 @@ function set_bc!(model; tracer::Symbol, pos::Symbol, bc_value::Union{Number, Abs
     @assert tracer in nut_names
 
     FT = model.FT
-    bc_value_d = bc_value
-    if isa(bc_value, AbstractArray)
-        bc_value_d = FT.(bc_value) |> array_type(model.arch)
+    bc_value_d = FT.(bc_value)
+    if isa(bc_value_d, AbstractArray)
+        bc_value_d = bc_value_d |> array_type(model.arch)
     end
     setproperty!(model.nutrients[tracer].bc, pos, bc_value_d)
     return nothing

src/Fields/halo_regions.jl

@@ -19,9 +19,9 @@
 @inline  fill_halo_north_vel!(c, H::Int, N::Int, ::Bounded)  = @views @. c[:, N+H+2:N+2H, :] = c[:, N+H+1:N+H+1, :]
 @inline fill_halo_bottom_vel!(c, H::Int, N::Int, ::Bounded)  = @views @. c[:, :, N+H+2:N+2H] = c[:, :, N+H+1:N+H+1]
 
-@inline   fill_halo_east_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[N+H+1:N+2H, :, :] = 0.0
-@inline  fill_halo_north_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[:, N+H+1:N+2H, :] = 0.0
-@inline fill_halo_bottom_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[:, :, N+H+1:N+2H] = 0.0
+@inline   fill_halo_east_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[N+H+1:N+2H, :, :] = 0.0f0
+@inline  fill_halo_north_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[:, N+H+1:N+2H, :] = 0.0f0
+@inline fill_halo_bottom_Gc!(c, H::Int, N::Int, ::Bounded) = @views @. c[:, :, N+H+1:N+2H] = 0.0f0
 
 fill_halo_east_vel!(c, H::Int, N::Int, TX::Periodic) = fill_halo_east!(c, H, N, TX)
 fill_halo_north_vel!(c, H::Int, N::Int, TY::Periodic) = fill_halo_north!(c, H, N, TY)