[llvm-commits] [llvm] r170509 - in /llvm/trunk/docs: Vectorizers.rst gcc-loops.png

Nadav Rotem nrotem at apple.com
Wed Dec 19 00:28:24 PST 2012


Author: nadav
Date: Wed Dec 19 02:28:24 2012
New Revision: 170509

URL: http://llvm.org/viewvc/llvm-project?rev=170509&view=rev
Log:
DOC: add a benchmarks that compares us to gcc and icc. 

Added:
    llvm/trunk/docs/gcc-loops.png   (with props)
Modified:
    llvm/trunk/docs/Vectorizers.rst

Modified: llvm/trunk/docs/Vectorizers.rst
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/docs/Vectorizers.rst?rev=170509&r1=170508&r2=170509&view=diff
==============================================================================
--- llvm/trunk/docs/Vectorizers.rst (original)
+++ llvm/trunk/docs/Vectorizers.rst Wed Dec 19 02:28:24 2012
@@ -188,6 +188,19 @@
 |fma  |trunc|nearbyint|
 +-----+-----+---------+
 
+Performance
+^^^^^^^^^^^
+
+This section shows the the execution time of Clang on a simple benchmark: 
+`gcc-loops <http://llvm.org/viewvc/llvm-project/test-suite/trunk/SingleSource/UnitTests/Vectorizer/>`._
+This benchmarks is a collection of loops from the GCC autovectorization 
+`page <http://gcc.gnu.org/projects/tree-ssa/vectorization.html>` by Dorit Nuzman._
+
+The chart below compares GCC-4.7, ICC-13, and Clang-SVN at -O3, running on a Sandybridge.
+The Y-axis shows time in msec. Lower is better.
+
+.. image:: gcc-loops.png
+
 The Basic Block Vectorizer
 ==========================
 

Added: llvm/trunk/docs/gcc-loops.png
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/docs/gcc-loops.png?rev=170509&view=auto
==============================================================================
Binary file - no diff available.

Propchange: llvm/trunk/docs/gcc-loops.png
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream





More information about the llvm-commits mailing list