Out of memory on device. To view more detail about available memory on the GPU, use 'gpuDevice()'

Hi, I am using a new pc and I have found this problem for the first time. The error is the following:

Error using gpuArray/subsref
Out of memory on device. To view more detail about available memory on the GPU, use 'gpuDevice()'. If the problem persists, reset the GPU by calling 'gpuDevice(1)'.

Error in matlab.internal.builtinhelper.circshift (line 87)
b = a(idx{:});

Error in ifftshift (line 27)
    x = circshift(x,ceil(size(x)/2));

Error in propagate2 (line 130)
    Eout_pad=ifft2(ifftshift(E0fft.*H.*mask));

Error in Prova24luglioOutofMemory (line 17)
D=propagate2(A, lambda, dz1, ps);

where A is a matrix, lambda, dz1 and ps numbers in order to propagate the matrix in free space.

If I execute gpuDevice()

  CUDADevice with properties:

                      Name: 'GeForce MX330'
                     Index: 1
         ComputeCapability: '6.1'
            SupportsDouble: 1
             DriverVersion: 11
            ToolkitVersion: 10.1000
        MaxThreadsPerBlock: 1024
          MaxShmemPerBlock: 49152
        MaxThreadBlockSize: [1024 1024 64]
               MaxGridSize: [2.1475e+09 65535 65535]
                 SIMDWidth: 32
               TotalMemory: 2.1475e+09
           AvailableMemory: 1.6409e+09
       MultiprocessorCount: 3
              ClockRateKHz: 1594000
               ComputeMode: 'Default'
      GPUOverlapsTransfers: 1
    KernelExecutionTimeout: 1
          CanMapHostMemory: 1
           DeviceSupported: 1
            DeviceSelected: 1
            

How can I overcome this problem?

I have found that if I have a size of the matrix A less than 4000 I have no problem but when I overcome that limit, I found this error message

[Matt J]

Well, it looks like you're out of memory. So, you have to use less of it (or get a graphics card with more memory).

[Alberto Paniate]

a graphic cards?

[Matt J]

Yes. The output of gpuDevice says that your GPU card (the GeForce MX330) has only about 2 GB RAM total. I don't know what other variables your propagate2() function is creating on the GPU but it is clearly exceeding the space you have there.