Search:

Type: Posts; User: Frizz

Search: Search took 0.00 seconds.

  1. Re: Runtime differences __global, __constant, __local

    > Yes, that is still necessary even if you pass the __local variable as a kernel argument.

    Uhh! So would the first function call in my kernel then by a barrier call? Like this here:

    __kernel...
  2. Re: Runtime differences __global, __constant, __local

    Thanks David.

    I am surprised that it is possible at all to use __local variables in a kernel function interface, since I've never seen that so far in sample code.

    What they usually do is...
  3. Runtime differences __global, __constant, __local

    Imagine I have some sort of filter algorithm.
    - in_array has the input data (vectorized, for faster access)
    - out_array gets the result of the filtering
    - filter is the filter itself.

    The...
  4. Thread: Cached Buffers

    by Frizz
    Replies
    2
    Views
    1,416

    Re: Cached Buffers

    To be more precise: Basically using texture caches, without the hassle of samplers, the same way as clEnqueueRead/Write/Buffer().

    Is this planned to come?
  5. Thread: Cached Buffers

    by Frizz
    Replies
    2
    Views
    1,416

    Cached Buffers

    Hi there,

    I was reading that the next version (1.2?) of OpenCL will support Cached Buffers. Is this true?
  6. Replies
    2
    Views
    1,182

    const variable / memory latency

    Hi,

    I have this kernel:

    __kernel void some_kernel(__global int* in_array, __global int* out_array)
    {
    uint tid = get_global_id(0);
    out_array[tid] = in_array[tid] *3;
    }
  7. Replies
    1
    Views
    1,887

    Apple FFT vs. AMD APP FFT

    Hi there,

    I would like to know if some of you already did a comparison of Apples widely known OpenCL FFT implementation versus AMD's new APP FFT implementation...
  8. Replies
    0
    Views
    1,427

    OpenCL : scope of application

    It seems that OpenCL is, at least at the moment, mostly used in scientific applications or image/movie editing programs.

    Are there any plans to speed up general purpose applications, like...
  9. Thread: Optimization

    by Frizz
    Replies
    1
    Views
    1,237

    Optimization

    Hi,

    I am just playing around with Apples OpenCL FFT code and added the following optimization:

    Instead of calculating "(dir*2.0f*M_PI*j/64)" over and over again, I cached the result in a...
Results 1 to 9 of 10