# Why can Flux not reduce this?

**URL:** <https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071>\
**Category:** GPU\
**Created:** [February 4, 2023, 11:11pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071 "2023-02-04T23:11:31Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Ahmed\_Salih](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/ahmed_salih/32/206579_2.png) [@Ahmed\_Salih](https://discourse.julialang.org/u/Ahmed_Salih)\
**Post date:** [February 4, 2023, 11:11pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/1 "2023-02-04T23:11:31Z")

</div>

Hello

I am trying:

```julia
using CUDA
S = rand(SVector{3,Float32},5)
DST = zeros(SVector{3,Float32},5)
I = [3,1,2,5,4]

# Works great on CPU

```

@CUDA.time NNlib.scatter!(+, DST,S,I)  
0.000004 seconds  
5-element Vector{SVector{3, Float32}}:  
[1.3667885, 1.081403, 1.1134366]  
[1.6543723, 0.50424564, 1.1448298]  
[0.8695842, 1.8623418, 1.398939]  
[1.0094697, 0.0052466393, 0.09184897]  
[1.9389011, 0.3291297, 1.2308507]

```julia

# Bugs out on GPU

 @CUDA.time NNlib.scatter!(+, CuArray(DST),CuArray(S),CuArray(I))
ERROR: InvalidIRError: compiling kernel #scatter_kernel!(typeof(+), CuDeviceVector{SVector{3, Float32}, 1}, CuDeviceVector{SVector{3, Float32}, 1}, CuDeviceVector{Int64, 1}) resulted in invalid LLVM IR
Reason: unsupported dynamic function invocation (call to atomic_cas!)

```

Anyone knows why?

This is an example, I need it to work on GPU for a more complex case

Kind regards

---

<div class="post-metadata">

**Author:** ![maleadt](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/maleadt/32/10097_2.png) [@maleadt](https://discourse.julialang.org/u/maleadt)\
**Post date:** [February 6, 2023, 9:21am UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/2 "2023-02-06T09:21:38Z")

</div>

That seems like a bug in NNlibCUDA.jl. Please file an issue there.

---

<div class="post-metadata">

**Author:** ![Ahmed\_Salih](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/ahmed_salih/32/206579_2.png) [@Ahmed\_Salih](https://discourse.julialang.org/u/Ahmed_Salih)\
**Post date:** [February 6, 2023, 6:05pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/3 "2023-02-06T18:05:44Z")

</div>

> [@maleadt](#):
>
> NNlibCUDA.jl

Thank you, I filed an issue 3 weeks ago:

> <https://github.com/FluxML/NNlib.jl/issues/507>
>
> Hi!
> 
> Using the package flux I want to scatter the following using NNlib:
> 
> 
> …\`\`\`julia
> using Flux
> 
> NNlib.scatter(+, \[SVector(1,1,1),SVector(1,1,1),SVector(1,1,1)\], \[3,1,2\])
> 
> 3-element Vector{SVector{3, Int64}}:
> \[1, 1, 1\]
> \[1, 1, 1\]
> \[1, 1, 1\]
> \`\`\`\`
> 
> Which works no problem. If I change the mid array to CuArray, then it works again, but tested that it is slow for large arrays (60k):
> 
> \`\`\`julia
> NNlib.scatter(+, CuArray(\[SVector(1,1,1),SVector(1,1,1),SVector(1,1,1)\]), \[3,1,2\])
> 
> 3-element CuArray{SVector{3, Int64}, 1, CUDA.Mem.DeviceBuffer}:
> \[1, 1, 1\]
> \[1, 1, 1\]
> \[1, 1, 1\]
> \`\`\`
> 
> If I try to do everything on GPU:
> 
> \`\`\`julia
> NNlib.scatter(+, CuArray(\[SVector(1,1,1),SVector(1,1,1),SVector(1,1,1)\]), CuArray(\[3,1,2\]))
> 
> ERROR: InvalidIRError: compiling kernel #scatter\_kernel!(typeof(+), CuDeviceVector{SVector{3, Int64}, 1}, CuDeviceVector{SVector{3, Int64}, 1}, CuDeviceVector{Int64, 1}) resulted in invalid LLVM IR
> Reason: unsupported dynamic function invocation (call to atomic\_cas!)
> \`\`\`
> 
> Which I think is an error?
> 
> More info: https://discourse.julialang.org/t/how-to-reduce-an-array/92945/14
> 
> Kind regards

I got an answer on there stating why it doesn’t work. Personally, I just thought “it should work” since StaticArrays is such a core part of Julia imo. One can get it to work doing some code like:

```julia
  for i = 1:3 #Length of static array element i.e. SVector{ **3** ,Float32}
        o = i - 1
        V_dst = @view reinterpret(eltype(eltype(dst)), dst)[begin+o:3:end]

        V_src_ = @view reinterpret(eltype(eltype(src)), src)[begin+o:3:end]
        V_src = @view V_src_[1:SYSTEM.MaxValidIndex[]]

        NNlib.scatter!(OP1,V_dst,V_src, @view(SYSTEM.I[1:SYSTEM.MaxValidIndex[]]))
        NNlib.scatter!(OP2,V_dst,V_src, @view(SYSTEM.J[1:SYSTEM.MaxValidIndex[]]))
    end

```

But I do not think this is the best fix, since I need to call NNlib.scatter!, 6 times to do this and possible lose out on some speed.

Kind regards

---

<div class="post-metadata">

**Author:** ![ToucheSir](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/touchesir/32/14411_2.png) [@ToucheSir](https://discourse.julialang.org/u/ToucheSir)\
**Post date:** [February 7, 2023, 10:40pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/4 "2023-02-07T22:40:24Z")

</div>

As I mentioned on that issue thread, you can turn this into 1 or 2 `scatter!` cols by using a 2D array instead.

---

<div class="post-metadata">

**Author:** ![Ahmed\_Salih](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/ahmed_salih/32/206579_2.png) [@Ahmed\_Salih](https://discourse.julialang.org/u/Ahmed_Salih)\
**Post date:** [February 8, 2023, 1:23pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/5 "2023-02-08T13:23:54Z")

</div>

Yes, but I really wish to stay using static arrays since it allows me for a relatively fast CPU execution as well. Do you happen to know if there is a way to “view” a vector of static arrays as a 2d matrix?

I only know how to get a single column out as shown above

Kind regards

---

<div class="post-metadata">

**Author:** ![Ahmed\_Salih](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/ahmed_salih/32/206579_2.png) [@Ahmed\_Salih](https://discourse.julialang.org/u/Ahmed_Salih)\
**Post date:** [February 8, 2023, 7:57pm UTC](https://discourse.julialang.org/t/why-can-flux-not-reduce-this/94071/6 "2023-02-08T19:57:55Z")

</div>

Got it to work, thank you