Search:

Type: Posts; User: vincentfpgarcia

Page 1 of 2 1 2

Search: Search took 0.00 seconds.

  1. I get that :) My question is different. The code...

    I get that :) My question is different. The code has been written by somebody else and there are like 10 layers of code above OpenCL.
    The way I see it is that the initial coder called a function...
  2. Replies
    3
    Views
    852

    So how can I modify a float4 in a function f() is...

    So how can I modify a float4 in a function f() is this float4 is a parameter of the function f() ?
  3. Replies
    3
    Views
    852

    Modify a float4 in a function

    Hi,

    I feel a bit stupid, I never had to deal with this simple problem in the past year.

    In my kernel I have two float4 variables : a and b.
    I want to call a function that will modify them.
    Is...
  4. The kernel does something like that: ...

    The kernel does something like that:



    __kernel void myKernel(
    __global const float * input,
    __global float * output,
    int size) {
    ...
    output[i] = something * input[i];
  5. Passing 2 times the same buffer as input ?

    Hi,

    I'm reading an already existing OpenCL code and it's the first time I see the following.
    The code as different layers of complexity so it's difficult for me to really understand what's going...
  6. Replies
    1
    Views
    821

    read_imagef : CL_RGBA vs CL_R

    Hi,

    Is it normal that reading a RGBA+FLOAT image is much slower than reading a R+FLOAT?
    In my case, it's about 3 times slower. I understand that in the first case I read 4 floats while I read...
  7. Replies
    4
    Views
    974

    Re: Precomputing array of 16 elements

    That's my guess too. Compilers are too smart nowadays :)
  8. Replies
    4
    Views
    974

    Re: Precomputing array of 16 elements

    Thank you for your answer. The code did compile though. Strange.
    Anyway, I tried the "__constant" version and it doesn't speed-up anything.
    It's like the more I code in OpenCL the less I...
  9. Replies
    4
    Views
    974

    Precomputing array of 16 elements

    Hi,

    Before I start, just want to say that (almost) everything works fine so it's just a question about what you think on this particular subject.
    Let's say that I have a 1D buffer of size N that...
  10. Re: Working with bits and perfomance issue

    As usual, it seems that what takes a lot of time is not what I thought it was.
    Le compiler was getting rid of some code when I was commenting other parts.
    Always difficult to understand where the...
  11. Re: Working with bits and perfomance issue

    It seems that the problem comes from the reading of crown.
    Indeed, if I change


    int descriptor = 0;
    for (int i=0; i<NB_SECTOR; ++i) {
    if (crown[i]>threshold){
    descriptor |=...
  12. Working with bits and perfomance issue

    Hi,

    I've some performance issues and I'd like your advice.
    I'm working with bits. It works fine but the following code takes about 95% of the computation time.
    First, I have an array of 16...
  13. Replies
    2
    Views
    866

    write_imagef and image1d

    Hi,

    So far, I was using an image2d to store a 1D look-up table. I've logically decided to put it into an image1d. However, my kernel cannot be build (CL_BUILD_PROGRAM_FAILURE). The following code...
  14. Re: Image Type : IMAGE2D vs IMAGE2D_ARRAY

    OK, got it and it's simple.
    It's possible in OpenCL to create an array of images instead of only one image.
    In this case, juste use CL_MEM_OBJECT_IMAGE2D_ARRAY.

    Thanks anyway.

    Vincent
  15. Image Type : IMAGE2D vs IMAGE2D_ARRAY

    Hi,

    I'm updating my OpenCL code and I'm trying to get rid of the old clCreateImage2D.
    The new version is clCreateImage and it requires different parameters.
    One of them is the image type. The...
  16. Re: Multiplication of 2 floats = 90% computation time?

    So yeap, the compiler skips the useless code. All the time is spent in reading the image (16 reads).
  17. Re: Multiplication of 2 floats = 90% computation time?

    Is the compiler smart enough to understand that I'm not using "L_xx" and "L_yy" so it skips all the unnecessary code? If yes, then I understand.
  18. Multiplication of 2 floats = 90% computation time?

    Hi,

    I have a kernel that I try to optimize and I have something strange. My kernel takes one image as input. For each pixel (c,r), it reads 12 values from the image in the neighborhood of (c,r)...
  19. Replies
    2
    Views
    808

    Re: Local extremum?

    Yeap, you were right, it's already pretty fast! It's definitely not the kernel I have to optimize...
    Thanks Clint.
  20. Replies
    2
    Views
    808

    Local extremum?

    Hi,

    I have a 2D image and I perform a classical local extremum (min-max) by reading the 8-neighborhood. My code is simple but I think absolutely not optimal for GPU.

    Is there a know algorithm...
  21. Replies
    7
    Views
    1,639

    Re: Simple question about clFinish()

    Yeap, thank you!
  22. Replies
    7
    Views
    1,639

    Re: Simple question about clFinish()

    If I understand what you are saying, if I use a simple command queue (i.e. I don't specify CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ ENABLE), then all what I put in the queue will be execute in order and one...
  23. Replies
    7
    Views
    1,639

    Re: Simple question about clFinish()

    I append data to the buffer using function g at the end of each step. I'm afraid that the blocking read at the end won't be enough but I'm not sure, hence my question.
  24. Replies
    7
    Views
    1,639

    Simple question about clFinish()

    Hi,

    Simple question related to the code bellow. It's not a real code, it doesn't compile, it's just a simple example.



    cl_mem image0;
    cl_mem image1;
    cl_mem image2;
    cl_mem buffer;
  25. Re: What's the deal with clEnqueueWriteBufferRect?

    Yeap, that's it, you're right. My solution was correct finally. Maybe the online documentation should be updated. Thanks for your help.
Results 1 to 25 of 47
Page 1 of 2 1 2