2.0.1 documentation updates (#1904)

* Updating release notes. * update notes. * update default behvaior of fdb ram cache in docs. * Removed statement about support for gfx8
ROCm · Aug 13, 2019 · 93074ff · 93074ff
1 parent 4e39a83
commit 93074ff
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 2 deletions.
diff --git a/doc/src/finddb.md b/doc/src/finddb.md
@@ -36,9 +36,9 @@ By default MIOpen will use the Find-Db. Users can disable the Find-Db by setting
 export MIOPEN_DEBUG_DISABLE_FIND_DB=1
 ```
 
-**Note:** The System Find-Db has the ability to be cached into memory and may increase performance dramatically. To enable this option use the cmake configuration flag:
+**Note:** The System Find-Db has the ability to be cached into memory and may increase performance dramatically. To disable this option use the cmake configuration flag:
 ```
--DMIOPEN_DEBUG_FIND_DB_CACHING=On
+-DMIOPEN_DEBUG_FIND_DB_CACHING=Off
 ```
 
 
diff --git a/doc/src/releasenotes.md b/doc/src/releasenotes.md
@@ -2,6 +2,26 @@
 ## MIOpen Release notes
 
 
+### 08/01/2019 [ 2.0.1 ]
+
+- This release contains bug fixes and performance improvements.
+- Additionally, the convolution algorithm Implicit GEMM is now enabled by default
+- Known issues: 
+    - Backward propagation for batch normalization in fp16 mode may trigger NaN in some cases
+    - Softmax Log mode may produce an incorrect result in back propagation
+
+Changes:
+
+- Added Winograd multi-pass convolution kernel
+- Fixed issue with hip compiler paths
+- Fixed immediate mode behavior with auto-tuning environment variable
+- Fixed issue with system find-db in-memory cache, the fix enable the cache by default
+- Improved logging
+- Improved how symbols are hidden in the library
+- Updated default behavior to enable implicit GEMM
+
+
+
 ### 07/08/2019 [ 2.0.0 ]
 
 - This release contains several new features including an immediate mode for selecting convolutions, bfloat16 support, new layers, modes, and algorithms.