openswr.github.io/perf.html at master · OpenSWR/openswr.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
<!DOCTYPE html>
<html>

  <head>
    <meta charset='utf-8'>
    <meta http-equiv="X-UA-Compatible" content="chrome=1">
    <meta name="description" content="OpenSWR : A High Performance, Highly Scalable software OpenGL implementation">

    <link rel="stylesheet" type="text/css" media="screen" href="stylesheets/stylesheet.css">

    <title>OpenSWR</title>
  </head>

  <body>

    <!-- HEADER -->
    <div id="header_wrap" class="outer">
        <header class="inner">
          <a id="mesa_banner" href="https://cgit.freedesktop.org/mesa/mesa/tree/src/gallium/drivers/swr">View source in Mesa</a>

                <a href="index.html">
                    <h1 id="project_title">OpenSWR</h1>
                </a>
                <a href="index.html">
                    <h2 id="project_tagline">A High Performance, Highly Scalable Software Rasterizer for OpenGL</h2>
                </a>

                <nav>
                    <ul class="navbar">
                        <li class="navbar"><a href="index.html">Overview</a></li>
                        <li class="navbar"><a href="index.html#build">Build</a></li>
                        <li class="navbar"><a href="gallery.html">Gallery</a></li>
                        <li class="navbar active"><a href="perf.html">Performance</a></li>
                        <li class="navbar" style="float:right"><a href="contact.html">Contact</a></li>
                    </ul>
                </nav>
        </header>
    </div>

    <!-- MAIN CONTENT -->
    <div id="main_content_wrap" class="outer">
        <section id="main_content" class="inner">
            <h3><a id="intro" class="anchor" href="#wintro" aria-hidden="true"><span class="octicon octicon-link"></span></a>
                Paraview and OpenSWR
            </h3>

<p>Here is a case study of ParaView, comparing OpenSWR performance to that of
the Mesa llvmpipe.</p>

<a href="jpg/OpenSWR_perf_vs_llvmpipe.png"><img src="jpg/OpenSWR_perf_vs_llvmpipe.png" width="95%"></a>

<p> In the top graph, OpenSWR outperforms llvmpipe by 29x in a scene with 2.9
million triangles and 51x when the geometry is increased to 106.2
million triangles.  </p>

<p> The bottom graph illustrates how OpenSWR performance scales with the number of
cores used for rendering.  </p>

System configuration:
<ul>
    <li>Intel<sup>&reg;</sup> Xeon<sup>&reg;</sup> E5-2699 v3 Processor, 2 x 18 cores, 2.3GHz
    <li>ParaView 4.3.1 (unmodified, latest released version)</li>
    <li>Mesa 10.5.1
    <li>OpenSWR "alpha 2"
    <li>TACC Isotropic turbulence data sets
    <ul>
        <li>Volumes ranging from 256<sup>3</sup> to 1280<sup>3</sup></li>
        <li>Isosurface + slice extracted from each volume producing geometries 2.9M to 106.2M triangles</li>
        <li>Model courtesy of the Texas Advanced Computing Center (TACC)
    </ul>
</ul>

    </section>
    </div>

    <!-- FOOTER  -->
    <div id="footer_wrap" class="outer">
      <footer class="inner">
        <p class="copyright">OpenSWR maintained by <a href="https://github.com/OpenSWR">OpenSWR</a>
        <br>For information about compiler optimizations, see our <a href="https://software.intel.com/en-us/articles/optimization-notice#opt-en">Optimization Notice</a>.
        <br>&copy;2009-2020 Intel Corporation
        <br><a href="https://www.intel.com/privacy">Privacy</a></p>
      </footer>
    </div>

  </body>
</html>