Search:

Type: Posts; User: Sarnath

Search: Search took 0.00 seconds.

  1. Replies
    4
    Views
    3,081

    Re: cl_mem and devices inside a context...

    pmdinu,

    You opened my eyes.. I did not realize that command queue is device specific... Hmm....

    Interesting... So, it is logically like dividing a buffer across multiple devices...

    Okay,...
  2. Replies
    4
    Views
    3,081

    Re: cl_mem and devices inside a context...

    pmdinu, Thanks for your reply.

    I still have some questions...

    Now, I understand that cl_mem buffer is associated with both devices...in the context. Fine.

    So, when I spawn a kernel, and if...
  3. Replies
    4
    Views
    3,081

    cl_mem and devices inside a context...

    I have a basic doubt about cl_mem thing..

    Say, I have a context made out of 2 CUDA GPUs. Now I create a cl_mem structure on top of it...

    Later, I issue a clEnqueueRead... request to read that...
  4. Replies
    8
    Views
    6,471

    Re: Array of pointers in device memory

    I think I differ here. CL_MEM can hold data... that is uni-dimensional. In CUDA, I could technically create an "int **a" on the device.... So, "a" points to an array of pointers each of which point...
  5. Replies
    8
    Views
    6,471

    Re: Array of pointers in device memory

    Not just array of pointers.... Even true heterogeneous computing (simultaneously harnessing CPU and GPU power) is NOT possible with this spec.
  6. Replies
    3
    Views
    3,571

    Re: True Heterogeneous Computing

    Oh great!

    We are working on a library close to that... But I will wait and see what vendors come up with.
  7. Replies
    8
    Views
    6,471

    Re: Array of pointers in device memory

    Thanks for the answer.

    Well, Is there a reason why you mention "1.0"?

    Is there any consideration for array of pointers?

    People often need to copy complex data-structures on to GPUs or other...
  8. Replies
    3
    Views
    3,571

    Re: True Heterogeneous Computing

    Thanks to Martin for suggesting me to post here.
  9. Replies
    8
    Views
    6,471

    Array of pointers in device memory

    Just wondering,

    How do I go about creating a cl_mem object which appears as an array of pointers to the kernel?

    I understand that a "cl_mem" argument appears as a "pointer" argument in the...
  10. Replies
    3
    Views
    3,571

    True Heterogeneous Computing

    I raised this issue in NVIDIA forums and one developer, at least, acknowledged this problem. I am just wondering what the OpenCL community has to say about this:

    "
    OpenCL spec says that it wants...
  11. Re: Is openCL suitable for task oriented parallelization really?

    The API for task parallelism do exist - Check EnqueueTask API
  12. Replies
    1
    Views
    4,306

    Re: When will OpenCL implementations arrive?

    NVIDIA has already demoed their OpenCL capable driver in SIGGRAPH ASIA. They have not officially released it though.
Results 1 to 12 of 12