Skip to content

Commit

Permalink
Merge pull request #44 from LLNL/release_v0.1
Browse files Browse the repository at this point in the history
Updating DiHydrogen to release v0.1
  • Loading branch information
bvanessen authored Sep 30, 2020
2 parents 9bca9d9 + ca4db3c commit 5025fd7
Show file tree
Hide file tree
Showing 2 changed files with 45 additions and 1 deletion.
2 changes: 1 addition & 1 deletion CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ cmake_minimum_required(VERSION 3.14.0)
cmake_policy(SET CMP0074 OLD)

project(DiHydrogen
VERSION 0.0.1
VERSION 0.1.0
DESCRIPTION "A distributed multilinear algebra interface library."
HOMEPAGE_URL "https://github.com/llnl/DiHydrogen"
LANGUAGES CXX)
Expand Down
44 changes: 44 additions & 0 deletions ReleaseNotes.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,44 @@
============================== Release Notes: v0.1 ==============================

Initial release of the DiHydrogen Distributed Tensor Library.

DiHydrogen is the second version of the Hydrogen fork of the
well-known distributed linear algebra library, Elemental. DiHydrogen
aims to be a basic distributed multilinear algebra interface with a
particular emphasis on the needs of the distributed machine learning
effort, LBANN.

API:

Performance optimizations:

Internal features:
- Support for the DistConv legacy library
- Added multi-dimensional tensor class
- Added support for older CUDA architectures v3.0+
- Use CUB for memory management
- Support for point-to-point (P2P) library
- Support for NVSHMEM halo exchange

DistConv Legacy:
- batchnorm
- channel-wise softmax
- concat
- convolution
- deconv
- leaky-relu
- pooling
- relu
- slice
- softmax

Build system:
- Require C++14
- Added support for cuDNN v8

Bug fixes:
- Fixed build issues with GCC8

Retired features:


0 comments on commit 5025fd7

Please sign in to comment.