Skip to content

Latest commit

 

History

History
32 lines (25 loc) · 1.24 KB

CHANGELOG.md

File metadata and controls

32 lines (25 loc) · 1.24 KB
Raw

Change log

[0.4.0] 10/02/2020

Changed

  • Split the memory management (CudaMatrix) from the CUBLAS invocation (CudaPipeline)
  • Moved all the allocation to the smart pointers inside CudaMatrix
  • Removed unused headers

[0.3.0] 26/09/2019

Added

  • Smart pointers to handle cuda resources
  • New CudaMatrix class
  • Use Eigen::MatrixXd
  • Check available memory in the GPU before computing

Removed

  • Template class, implementation only for double available
  • Triple tensor product
  • Shapes struct

[0.2.0] 27/08/2019

Added

  • Tensor matrix multiplacation using gemmbatched.
  • Async calls to memory copies.
  • Properly free memory after the tensor operation is done.

[0.1.0]

New

  • Use a template function to perform matrix matrix multiplacation using CUBLAS.
  • Use either pinned (default) or pageable memory, see cuda optimizations.