Search:

Type: Posts; User: santtt

Search: Search took 0.00 seconds.

  1. Replies
    1
    Views
    1,554

    Problem with code running on CPU

    Well, i need to run a code on a C2D T7250, but the CL_DEVICE_NOT_FOUND flag is turned on when i run it. I try to compile the code with de ATI STREAM OpenCL library, but the error still remains. The...
  2. Replies
    2
    Views
    1,874

    Re: How can I simulate double precision?

    See section 9.3 of the specification document... basically put the macro #pragma OPENCL EXTENSION cl_khr_fp64: enable and then normally, declare as cl_double, cl_double2, ... remember that they are...
  3. Replies
    1
    Views
    1,893

    Re: Initializing constant variables

    You can create the constants variables on the host, then initialize a memory object that contains it, do the corresponding copy of data, add this memory object in the kernel args and finally add the...
  4. Replies
    14
    Views
    9,141

    Re: opencl build error..!!

    I got an error like yours and the answer was casting types. Cast everything you see and compile again!
  5. Replies
    5
    Views
    4,626

    Re: How to perform atomic sums on floats

    No.... you don't understand me. let's put an example. You have one piece of memory for the result value, and a piece of memory with some values. When you have to sum all this values and return a...
  6. Re: Problem with moving an array pointer in the Kernel

    Well in this code it seems that you are concatenating height times the first w2 values of buff, in the middle you do buff += pitch but buff is an array and pitch is a value, this is not possible........
  7. Replies
    14
    Views
    9,141

    Re: opencl build error..!!

    sorry, mfc? what is that?
  8. Replies
    5
    Views
    4,626

    Re: How to perform atomic sums on floats

    The code is not right... and does not perform summation that you want. The problem is in te way that you perform de operation, you are telling to OpenCL that every workItem update the same portion of...
  9. Replies
    14
    Views
    9,141

    Re: opencl build error..!!

    Do you have multiples kernels on the opencl kernel file? If that, try to make a file with only one kernel.
  10. Replies
    14
    Views
    9,141

    Re: opencl build error..!!

    It seems that you have an kernel code error, i suggest you that read de code in the line that appear in the error and look for an error in the neighborhood.
  11. Re: Effective gain by using padding on workgroup size

    Theorically yes, because the size tends to be in proportion with the warp, more preciselly, the walf-warp. I recommend you to read the OpenCL Programming Guide, and the others related to learn about...
  12. Replies
    14
    Views
    9,141

    Re: opencl build error..!!

    Are you sure your openCL kernel it's OK? Try commenting the code of the kernel and seeing if compile. If it happens, there is a kernel code error, with CL_PROGRAM_BUILD_STATUS flag in...
  13. Replies
    1
    Views
    1,697

    Re: Execute opencl kernel problem..

    Can you put some lines of your code? I don't undestand you, and maybe the code helps me to helps you :) .
  14. Re: Does the number of arguments affect to the performance ?

    Well, in my opinion the number is not the problem, the problem is p.e. the size of the elements that you are passing throw the kernel, be carefully in the size of allocations in local memory. If the...
  15. Re: Size of a structure. Should it be multiple of 16 ?

    Well the aligment multiple by 16 comes by the way the SIMT pushes the work on the computeElements. There exists some called HalfWarp, that consist on 16 workItems that have a determinated size in...
  16. Replies
    4
    Views
    1,949

    Re: How to assure kernel execution ended?

    In C i use profiling for timming my kernel, you have to past to the queue properties the CL_QUEUE_PROFILING_ENABLE flag and then take de event object from clEnqueueNDRangeKernel. After that put that...
  17. Re: Poor bandwith on matrix multiplication with local memory?

    Thanks for your reply, the maps were only for play and learn how is it works. In my original code i use MEM_COPY_HOST_PTR with a valid pointer where is the info for passing the data to the device....
  18. Re: Poor bandwith on matrix multiplication with local memory?

    im reading the papers, but still no idea for the poor performance... im trying to implement another method, but at the time im confused...
  19. Poor bandwith on matrix multiplication with local memory?

    Well, i'm trying to do the best i can to increase BW on matrix multiplication but my efforts aren't coming to where i want... in a Tesla C1060 the operation of multiplying two matrix of 2048x2048 is...
  20. Appear to be just few results when i execute my OpenCL code?

    I'm doing a .cpp file that contains linear algebra functions, but i'm having a problem when the results are printed. Always a little part is well done, at the beginning, and the rest is zero. Where...
  21. Replies
    4
    Views
    2,323

    Re: Troubles with clBuildProgram...

    Thanks for your answers, i'm looking the clGetProgramBuildInfo and it appear to be an error on the kernel2.cl, fantastic!. I will continue looking this code this night and tomorrow will tells you if...
  22. Replies
    4
    Views
    2,323

    Troubles with clBuildProgram...

    First hello to everybody and sorry for my english i'm an spanish speaker =), its my first time here and i want to congrats to everyone that colaborate ont his subject. Let's go to the problem, i...
Results 1 to 22 of 23