[llvm-commits] CVS: llvm-www/pubs/2009-02-PPoPP-MappingParallelism.html 2009-02-PPoPP-MappingParallelism.pdf pubs.js

Sat Jun 27 00:21:13 PDT 2009

Changes in directory llvm-www/pubs:

2009-02-PPoPP-MappingParallelism.html added (r1.1)
2009-02-PPoPP-MappingParallelism.pdf added (r1.1)
pubs.js updated: 1.21 -> 1.22
---
Log message:

Add "Mapping parallelism to multi-cores: a machine learning based approach"
from PPoPP'09


---
Diffs of the changes:  (+70 -0)

 2009-02-PPoPP-MappingParallelism.html |   63 ++++++++++++++++++++++++++++++++++
 2009-02-PPoPP-MappingParallelism.pdf  |    0 
 pubs.js                               |    7 +++
 3 files changed, 70 insertions(+)


Index: llvm-www/pubs/2009-02-PPoPP-MappingParallelism.html
diff -c /dev/null llvm-www/pubs/2009-02-PPoPP-MappingParallelism.html:1.1
*** /dev/null	Sat Jun 27 02:20:28 2009

--- llvm-www/pubs/2009-02-PPoPP-MappingParallelism.html	Sat Jun 27 02:20:18 2009
***************
*** 0 ****
--- 1,63 ----
+ <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
+ <html>
+ <head>
+   <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
+   <link rel="stylesheet" href="../llvm.css" type="text/css" media="screen">
+   <title>Mapping parallelism to multi-cores: a machine learning based approach</title>
+ </head>
+ <body>
+ 
+ <div class="pub_title">
+ Mapping parallelism to multi-cores: a machine learning based approach
+ </div>
+ <div class="pub_author">
+   Zheng Wang and Michael F.P. O'Boyle
+ </div>
+ 
+ <h2>Abstract:</h2>
+ <blockquote>
+ The efficient mapping of program parallelism to multi-core processors is highly dependent on the underlying architecture. This paper proposes a portable and automatic compiler-based approach to mapping such parallelism using machine learning. It develops two predictors: a data sensitive and a data insensitive predictor to select the best mapping for parallel programs. They predict the number of threads and the scheduling policy for any given program using a model learnt off-line. By using low-cost profiling runs, they predict the mapping for a new unseen program across multiple input data sets. We evaluate our approach by selecting parallelism mapping configurations for OpenMP programs on two representative but different multi-core platforms (the Intel Xeon and the Cell processors). Performance of our technique is stable across programs and architectures. On average, it delivers above 96% performance of the maximum available on both platforms. It achieve, on average, a 37%!
  (up to 17.5 times) performance improvement over the OpenMP runtime default scheme on the Cell platform. Compared to two recent prediction models, our predictors achieve better performance with a significant lower profiling cost.
+ </blockquote>
+ 
+ <h2>Published:</h2>
+ <blockquote>
+ "Mapping parallelism to multi-cores: a machine learning based approach"<br>
+ Zheng Wang and Michael F.P. O'Boyle.<br>
+ <i>
+ Proceedings of the 14th ACM SIGPLAN symposium on Principles and practice of parallel programming (PPoPP'09)
+ </i>, Raleigh, NC, USA, February 2009.
+ </blockquote>
+ 
+ <h2>Download:</h2>
+ <h3>Paper:</h3>
+ <ul>
+   <li><a href="2009-02-PPoPP-MappingParallelism.pdf">
+   Mapping parallelism to multi-cores: a machine learning based approach
+   </a> (PDF)</li>
+ </ul>
+ 
+ <h2>BibTeX Entry:</h2>
+ <pre>
+ @inproceedings{1504189,
+  author = {Wang, Zheng and O'Boyle, Michael F.P.},
+  title = {Mapping parallelism to multi-cores: a machine learning based approach},
+  booktitle = {PPoPP '09: Proceedings of the 14th ACM SIGPLAN symposium on Principles and practice of parallel programming},
+  year = {2009},
+  isbn = {978-1-60558-397-6},
+  pages = {75--84},
+  location = {Raleigh, NC, USA},
+  doi = {http://doi.acm.org/10.1145/1504176.1504189},
+  publisher = {ACM},
+  address = {New York, NY, USA},
+  }
+ </pre>
+ 
+ <!-- *********************************************************************** -->
+ <hr>
+   <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
+   src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
+   <a href="http://validator.w3.org/check/referer"><img
+   src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
+ 
+ </body>
+ </html>


Index: llvm-www/pubs/2009-02-PPoPP-MappingParallelism.pdf


Index: llvm-www/pubs/pubs.js
diff -u llvm-www/pubs/pubs.js:1.21 llvm-www/pubs/pubs.js:1.22
--- llvm-www/pubs/pubs.js:1.21	Sat Jun 27 02:06:44 2009
+++ llvm-www/pubs/pubs.js	Sat Jun 27 02:20:17 2009
@@ -86,6 +86,13 @@
    month: 3,
    year: 2009},
 
+   {url: '2009-02-PPoPP-MappingParallelism.html',
+    title: 'Mapping parallelism to multi-cores: a machine learning based approach',
+    author: "Zheng Wang and Michael F.P. O'Boyle",
+    published: "Proc. of the 14th ACM SIGPLAN symposium on Principles and practice of parallel programming (PPoPP'09)",
+    month: 2,
+    year: 2009},
+
    {url: '2009-01-VMCAI-ScalableMemoryModel.html',
     title: 'A Scalable Memory Model for Low-Level Code',
     author: 'Zvonimir Rakamaric and Alan J. Hu',




