Search:

Type: Posts; User: roger512

Search: Search took 0.00 seconds.

  1. Replies
    1
    Views
    1,629

    adding some cool features like : - function...

    adding some cool features like :

    - function pointers
    - interlaced switch/loop could be usefull too.
  2. Replies
    2
    Views
    1,173

    Hello, the compiler will prolly factorize...

    Hello,



    the compiler will prolly factorize it that way :

    barrier(CLK_LOCAL_MEM_FENCE | CLK_GLOBAL_MEM_FENCE)

    if (<condition>) {
    ...
  3. Replies
    4
    Views
    1,691

    Doesn't work, i tried to put it at the begining...

    Doesn't work, i tried to put it at the begining and the end it does the same thing.
  4. Replies
    4
    Views
    1,691

    make sense, I don't know why your talking...

    make sense,

    I don't know why your talking about cudaMemcpyTosymbol() here, it's ptx from opencl code.

    It's a glitch in nvidia opencl implementation, my code works with ATI cards.
  5. Replies
    1
    Views
    1,629

    goto statement

    Hello,

    This can seems an odd request but i do think it would offer new possibilities to OpenCL if the goto statement (for a warp/wavefront) was available. Let me try to explain.

    1) It would...
  6. Replies
    4
    Views
    1,691

    constant memory issue

    hi, i m getting a very strange issue with constant memory lately.

    My code work fine if a int buffer ( 2 elements ) is in global mem (__global) but when i swap it to (__constant) the result become...
  7. Replies
    2
    Views
    1,453

    Re: register usage : float3 vs float4

    ok it seems NVIDIA compiler is able to save register when i use float3 instead of float4.
    And i can keep float4 in global to keep the alignment using as_float3 convertor.

    Roger
  8. Replies
    2
    Views
    1,453

    register usage : float3 vs float4

    Hi everybody,

    I'm currently working with opencl and i'm getting issues with a high amount of registers per thread in my main kernel.

    The main kernel use a quite large amount of float4 but...
Results 1 to 8 of 9