<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
          "http://www.w3.org/TR/html4/strict.dtd">
<!-- Material used from: HTML 4.01 specs: http://www.w3.org/TR/html401/ -->
<html>
<head> <META http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
  <title>Polly - Performance</title>
  <link type="text/css" rel="stylesheet" href="menu.css">
  <link type="text/css" rel="stylesheet" href="content.css">
</head>
<body>
<div id="box">
<!--#include virtual="menu.html.incl"-->
<div id="content">
<h1>Performance</h1>

<p>To evaluate the performance benefits Polly currently provides we compiled the
<a href="https://web.cse.ohio-state.edu/~pouchet.2/software/polybench/">Polybench
2.0</a> benchmark suite.  Each benchmark was run with double precision floating
point values on an Intel Core Xeon X5670 CPU @ 2.93GHz (12 cores, 24 thread)
system. We used <a href="https://sourceforge.net/projects/pocc/files/">PoCC</a> and the included <a
href="http://pluto-compiler.sf.net">Pluto</a> transformations to optimize the
code. The source code of Polly and LLVM/clang was checked out on
25/03/2011.</p>

<p>The results shown were created fully automatically without manual
interaction. We did not yet spend any time to tune the results. Hence
further improvements may be achieved by tuning the code generated by Polly, the
heuristics used by Pluto or by investigating if more code could be optimized.
As Pluto was never used at such a low level, its heuristics are probably
far from perfect. Another area where we expect larger performance improvements
is the SIMD vector code generation. At the moment, it rarely yields to
performance improvements, as we did not yet include vectorization in our
heuristics. By changing this we should be able to significantly increase the
number of test cases that show improvements.</p>

<p>The polybench test suite contains computation kernels from linear algebra
routines, stencil computations, image processing and data mining. Polly
recognizes the majority of them and is able to show good speedup. However,
to show similar speedup on larger examples like the SPEC CPU benchmarks Polly
still misses support for integer casts, variable-sized multi-dimensional arrays
and probably several other constructs. This support is necessary as such
constructs appear in larger programs, but not in our limited test suite.

<h2> Sequential runs</h2>

For the sequential runs we used Polly to create a program structure that is
optimized for data-locality. One of the major optimizations performed is tiling.
The speedups shown are without the use of any multi-core parallelism. No
additional hardware is used, but the single available core is used more
efficiently.
<h3> Small data size</h3>
<img src="images/performance/sequential-small.png" /><br />
<h3> Large data size</h3>
<img src="images/performance/sequential-large.png" />
<h2> Parallel runs</h2>
For the parallel runs we used Polly to expose parallelism and to add calls to an
OpenMP runtime library. With OpenMP we can use all 12 hardware cores
instead of the single core that was used before. We can see that in several
cases we obtain more than linear speedup. This additional speedup is due to
improved data-locality.
<h3> Small data size</h3>
<img src="images/performance/parallel-small.png" /><br />
<h3> Large data size</h3>
<img src="images/performance/parallel-large.png" />
</div>
</div>
</body>
</html>