Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improve Batched GEMV Speed #113

Open
wants to merge 20 commits into
base: master
Choose a base branch
from
Open

Improve Batched GEMV Speed #113

wants to merge 20 commits into from

Commits on Jul 11, 2016

  1. Refactor gemv_prog Geometry into a separate class

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    74fb5c2 View commit details
    Browse the repository at this point in the history
  2. Compute bandwidth and FLOPS in test_clra_gemv

    In addition, run GEMV multiple times to improve the accuracy of the result.
    
    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    e5e445d View commit details
    Browse the repository at this point in the history
  3. Print profiling stats when profiling is not 0

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    d4f5aa1 View commit details
    Browse the repository at this point in the history
  4. Show profiling runtime to five decimal places

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    a496cf3 View commit details
    Browse the repository at this point in the history
  5. Use list instead of map when generating profiling columns

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    a982813 View commit details
    Browse the repository at this point in the history
  6. Support both row and column-major RaggedArray

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    e8c0759 View commit details
    Browse the repository at this point in the history
  7. Descriptive block_impl NotImplementedError messages

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    162d7eb View commit details
    Browse the repository at this point in the history
  8. Add some comments to geometry for clarification

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    621f776 View commit details
    Browse the repository at this point in the history
  9. Generate a_s0 or a_s1 in cl_geometry_and_textconf

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    6690d23 View commit details
    Browse the repository at this point in the history
  10. Make clra_gemv reduce_impl column-major

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    3e97b6a View commit details
    Browse the repository at this point in the history
  11. Make clra_gemv block_impl column-major

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 11, 2016
    Configuration menu
    Copy the full SHA
    7a1b166 View commit details
    Browse the repository at this point in the history

Commits on Jul 16, 2016

  1. Add clra_gemv one_thread_per_row_impl

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 16, 2016
    Configuration menu
    Copy the full SHA
    1308e04 View commit details
    Browse the repository at this point in the history
  2. Fix RaggedArray stride and order assert

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 16, 2016
    Configuration menu
    Copy the full SHA
    2c8d10f View commit details
    Browse the repository at this point in the history

Commits on Jul 31, 2016

  1. Implement current LIF model in nengo

    This commit implements the more accurate LIF model, implemented in Nengo
    2.1.1, in OpenCL.
    
    A new boolean argument `fastlif' is also added in plan_lif, which
    defaults to False.
    
    See <nengo/nengo#975> for details regarding the
    new LIF model.
    
    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Jul 31, 2016
    Configuration menu
    Copy the full SHA
    43d747b View commit details
    Browse the repository at this point in the history

Commits on Aug 1, 2016

  1. Modify clra_nonlineralities to accept column-major

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    161a032 View commit details
    Browse the repository at this point in the history
  2. Fix CLRaggedArray __setitem__ for column-major

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    92349e4 View commit details
    Browse the repository at this point in the history
  3. Use column-major RaggedArray by default

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    e0c3297 View commit details
    Browse the repository at this point in the history
  4. Add plan_pretuned_gemv

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    0c3b028 View commit details
    Browse the repository at this point in the history
  5. Fix column-major view strides

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    2a33c21 View commit details
    Browse the repository at this point in the history
  6. Use plan_one_thread_per_row_gemv in Simulator

    Signed-off-by: Shaun Ren <[email protected]>
    shaunren committed Aug 1, 2016
    Configuration menu
    Copy the full SHA
    d4fa2a4 View commit details
    Browse the repository at this point in the history