summaryrefslogtreecommitdiff
path: root/CHANGELOG
diff options
context:
space:
mode:
authorCNugteren <web@cedricnugteren.nl>2015-07-24 08:16:41 +0200
committerCNugteren <web@cedricnugteren.nl>2015-07-24 08:16:41 +0200
commita76dc2f09c283d6ee8d92e1c4ababc6a7cb22c91 (patch)
tree6dcf2662557283ac89e314f0e56b2cf267b0fc6f /CHANGELOG
parent547b7afffce27cc134e069f2fe64869ea0977acd (diff)
Updated the docs to reflect the performance improvements
Diffstat (limited to 'CHANGELOG')
-rw-r--r--CHANGELOG3
1 files changed, 2 insertions, 1 deletions
diff --git a/CHANGELOG b/CHANGELOG
index c6cf612a..0fee63af 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -1,7 +1,8 @@
Development version (next release)
- Re-organized test/client infrastructure to avoid code duplication
-- Bypasses pre/post-processing kernels if possible (in level-3 routines)
+- Added an optional bypass for pre/post-processing kernels in level-3 routines
+- Significantly improved performance of level-3 routines on AMD GPUs
- Added level-3 routines:
* CHEMM/ZHEMM
* SSYRK/DSYRK/CSYRK/ZSYRK