Unable to upgrade to the latest version

[gzhang]

(@v1.5) pkg> up CUDA
Updating registry at C:\Users\gzhang\.julia\registries\General
Updating git-repo https://github.com/JuliaRegistries/General.git
Installed IJulia ──────────── v1.21.4
Installed DiffEqSensitivity ─ v6.32.0
No Changes to C:\Users\gzhang\.julia\environments\v1.5\Project.toml
Updating C:\Users\gzhang\.julia\environments\v1.5\Manifest.toml
[41bf760c] ↑ DiffEqSensitivity v6.31.6 ⇒ v6.32.0
[7073ff75] ↑ IJulia v1.21.3 ⇒ v1.21.4
Building IJulia → C:\Users\gzhang\.julia\packages\IJulia\rWZ9e\deps\build.log

(@v1.5) pkg> st CUDA
Status C:\Users\gzhang\.julia\environments\v1.5\Project.toml
[052768ef] CUDA v0.1.0

(@v1.5) pkg>

system info
win 10 64
CUDA 11.0
Julia 1.5.2

[maleadt]

Not a bug. Some package is holding back CUDA.jl, try add CUDA@1.3.3 to see why.

For support questions, please use Discourse or Slack, not the bug tracker.

[gzhang]

@maleadt OK, ths. it is fixed.

(@v1.5) pkg> add CUDA@1.3.3
Resolving package versions...
Installed Lathe ──────────── v0.0.9
Installed Juno ───────────── v0.8.4
Installed NeuralPDE ──────── v2.3.0
Installed DiffEqFlux ─────── v1.23.0
Installed Atom ───────────── v0.12.23
Installed Optim ──────────── v1.2.0
Installed Zygote ─────────── v0.5.8
Installed ModelingToolkit ── v3.20.1
Installed Flux ───────────── v0.11.1
Installed CSV ────────────── v0.5.23
Installed ArrayFire ──────── v1.0.7
Installed IndexedTables ──── v0.12.4
Installed CUDA ───────────── v1.3.3
Installed GalacticOptim ──── v0.2.2
Installed GPUCompiler ────── v0.6.1
Installed NNlib ──────────── v0.7.4
Installed ChainRules ─────── v0.7.21
Installed JLD2 ───────────── v0.2.1
Installed ChainRulesCore ─── v0.9.10
Installed JuliaFormatter ─── v0.8.2
Installed Knet ───────────── v1.4.2
Installed DistributionsAD ── v0.6.9
Installed Compat ─────────── v3.17.0
Installed Adapt ──────────── v2.3.0
Installed TSVD ───────────── v0.4.1
Installed GPUArrays ──────── v5.2.1
Installed JuliaInterpreter ─ v0.8.1
Updating C:\Users\gzhang\.julia\environments\v1.5\Project.toml
[b19378d9] ↑ ArrayFire v1.0.6 ⇒ v1.0.7
[c52e3926] ↑ Atom v0.12.21 ⇒ v0.12.23
[336ed68f] ↓ CSV v0.7.7 ⇒ v0.5.23
[052768ef] ↑ CUDA v0.1.0 ⇒ v1.3.3
[aae7a2af] ↑ DiffEqFlux v1.17.0 ⇒ v1.23.0
[587475ba] ↑ Flux v0.10.4 ⇒ v0.11.1
[a93385a2] ↓ JuliaDB v0.13.0 ⇒ v0.12.0
[e5e0dc1b] ↑ Juno v0.8.3 ⇒ v0.8.4
[1902f260] ↑ Knet v1.3.7 ⇒ v1.4.2
[38d8eb38] ↓ Lathe v0.1.1 ⇒ v0.0.9
[961ee093] ↑ ModelingToolkit v3.20.0 ⇒ v3.20.1
[315f7962] ↑ NeuralPDE v2.0.0 ⇒ v2.3.0
[a15396b6] ↓ OnlineStats v1.5.5 ⇒ v0.23.0
[429524aa] ↑ Optim v0.20.1 ⇒ v1.2.0
[e88e6eb3] ↑ Zygote v0.4.22 ⇒ v0.5.8
Updating C:\Users\gzhang\.julia\environments\v1.5\Manifest.toml
[79e6a3ab] ↑ Adapt v1.1.0 ⇒ v2.3.0
[b19378d9] ↑ ArrayFire v1.0.6 ⇒ v1.0.7
[c52e3926] ↑ Atom v0.12.21 ⇒ v0.12.23
[fa961155] ↑ CEnum v0.2.0 ⇒ v0.4.1
[336ed68f] ↓ CSV v0.7.7 ⇒ v0.5.23
[052768ef] ↑ CUDA v0.1.0 ⇒ v1.3.3
[3895d2a7] - CUDAapi v4.0.0
[c5f51814] - CUDAdrv v6.3.0
[be33ccc6] - CUDAnative v3.0.4
[49dc2e85] - Calculus v0.5.1
[324d7699] ↓ CategoricalArrays v0.8.3 ⇒ v0.7.7
[082447d4] ↑ ChainRules v0.6.5 ⇒ v0.7.21
[d360d2e6] ↑ ChainRulesCore v0.8.1 ⇒ v0.9.10
[34da2185] ↑ Compat v2.2.1 ⇒ v3.17.0
[f68482b8] - Cthulhu v1.2.2
[3a865a2d] - CuArrays v2.2.2
[a93c6f00] ↓ DataFrames v0.21.7 ⇒ v0.20.0
[01453d9d] - DiffEqDiffTools v1.7.0
[aae7a2af] ↑ DiffEqFlux v1.17.0 ⇒ v1.23.0
[ced4e74d] + DistributionsAD v0.6.9
[48062228] + FilePathsBase v0.7.0
[587475ba] ↑ Flux v0.10.4 ⇒ v0.11.1
[1eca21be] - FoldingTrees v1.0.1
[d9f16b24] + Functors v0.1.0
[0c68f7d7] ↑ GPUArrays v3.4.1 ⇒ v5.2.1
[61eb1bfa] ↑ GPUCompiler v0.3.0 ⇒ v0.6.1
[a75be94c] + GalacticOptim v0.2.2
[6deec6e2] ↓ IndexedTables v0.12.6 ⇒ v0.12.4
[033835bb] ↑ JLD2 v0.1.14 ⇒ v0.2.1
[a93385a2] ↓ JuliaDB v0.13.0 ⇒ v0.12.0
[98e50ef6] ↑ JuliaFormatter v0.7.6 ⇒ v0.8.2
[aa1ae85d] ↑ JuliaInterpreter v0.7.26 ⇒ v0.8.1
[e5e0dc1b] ↑ Juno v0.8.3 ⇒ v0.8.4
[1902f260] ↑ Knet v1.3.7 ⇒ v1.4.2
[929cbde3] ↑ LLVM v1.7.0 ⇒ v2.0.0
[38d8eb38] ↓ Lathe v0.1.1 ⇒ v0.0.9
[7f8f8fb0] + LearnBase v0.2.2
[30fc2ffe] + LossFunctions v0.5.1
[961ee093] ↑ ModelingToolkit v3.20.0 ⇒ v3.20.1
[d41bc354] ↑ NLSolversBase v7.5.0 ⇒ v7.7.0
[872c559c] ↑ NNlib v0.6.6 ⇒ v0.7.4
[315f7962] ↑ NeuralPDE v2.0.0 ⇒ v2.3.0
[4d1e1d77] ↑ Nullables v0.0.8 ⇒ v1.0.0
[a15396b6] ↓ OnlineStats v1.5.5 ⇒ v0.23.0
[925886fa] ↓ OnlineStatsBase v1.4.4 ⇒ v0.10.2
[429524aa] ↑ Optim v0.20.1 ⇒ v1.2.0
[69de0a69] ↓ Parsers v1.0.10 ⇒ v0.3.12
[06bb1623] + PenaltyFunctions v0.1.2
[91c51154] - SentinelArrays v1.2.15
[09ab397b] ↓ StructArrays v0.4.4 ⇒ v0.4.1
[856f2bd8] - StructTypes v1.1.0
[7522ee7d] ↓ SweepOperator v0.3.0 ⇒ v0.2.0
[9449cd9e] ↑ TSVD v0.4.0 ⇒ v0.4.1
[bd369af6] ↓ Tables v1.0.5 ⇒ v0.2.11
[ea10d353] ↓ WeakRefStrings v0.6.2 ⇒ v0.5.8
[e88e6eb3] ↑ Zygote v0.4.22 ⇒ v0.5.8

[maleadt]

Great. Some comments:

• please use triple backticks when posting output like that
• don't have CUDA.jl co-installed with CuArrays/CUDAnative/CUDAdrv/CUDAapi (it got removed here, but that was likely what was holding CUDA.jl back)