Calculating kmeans on a GPU

Kind time of the day, dear inhabitants of the forum. I want to ask whether anyone can help me calculate the clustering of kmeans on a GPU it is not included in the standard function gpuArray but it can be calculated as a user-defined function But how to do that? I'm new to matlab and can not figure out how to implement it.

Answer 1

[Joss Knight]

gpuArray support for kmeans was added to MATLAB in R2016a

>> help gpuArray/kmeans
 kmeans K-means clustering for GPU data
    IDX = kmeans(X, K)
    [IDX, C] = kmeans(X, K)
    [IDX, C, SUMD] = kmeans(X, K)
    [IDX, C, SUMD, D] = kmeans(X, K)
    [ ... ] = kmeans(..., 'PARAM1',val1, 'PARAM2',val2, ...)

      Example:
          N = 1e6;
          X = gpuArray([randn(N,2)+ones(N,2); randn(N,2)-ones(N,2)]);
          [cidx, ctrs] = kmeans(X, 2);

      See also kmeans, gpuArray.

Answer 2

[Walter Roberson]

https://github.com/NVIDIA/kmeans

https://blog.sourced.tech/post/towards_kmeans_on_gpu/

Answer 3

[Mailz Daizan]

Thank you all very much for the answers, I'll get to the weekend before PC I'll try it!