{{{
#!html

<h2>
Old Hardware Benchmarks Running UCSF Chimera
</h2>

<p>
There are two types of benchmarks: the volume benchmarks and the molecule benchmarks.

<h3>Volume benchmarks</h3>

<p>
The volume scores (Surface, Mesh, Contour, Solid, Recolor)
give the edge size of a cube shaped volume data array such that
rendering or contouring or recoloring can be performed 10 times per second.
Data of this size can be easily manipulated in Chimera.
</p>

<p>
Doubling a surface or mesh rendering score means rendering 4 times as
many triangles or lines per second.  Doubling a contouring, solid rendering
or solid recoloring score means processing 8 times as many data points
per second.
</p>

<h3>Molecule benchmarks</h3>
<p>
The molecule scores (Null, Wire, Stick, BStick, Ribbon, Sphere, Ops)
give the frames-per-second for a given molecule and representation.
(See the 
<a href="http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/benchmark/benchmark.html">benchmark extension</a>
documentation for details.)


<h3>Benchmark Extension</h3>

<p>
The
<a href="http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/benchmark/benchmark.html">benchmark extension</a>
can be started from the Tools/Utilities/Benchmark menu entry.
If you send us 
(<a href="mailto:chimera-users@cgl.ucsf.edu">chimera-users@cgl.ucsf.edu</a>)
your scores and machine information, we will include them in the table below.
</p>


<h3>
Scores for Chimera versions 1.2509 - 1.2524 (May 5 - June 5, 2008).
</h3>


<table border=1 style="background-color: #F0F0F0; border-spacing: 0">

<tr><th><th><th><th colspan=5>Volume Scores<th colspan=6>Molecule Scores

<tr>
<th>Graphics
<th>Graphics Memory (Mb)

<th>OS Type
<th>Surface
<th>Mesh
<th>Contour
<th>Solid
<th>Recolor
<th>Wire
<th>Stick
<th>BStick
<th>Ribbon
<th>Sphere
<th>Ops
<th>Notes
<th>Test Date
<th>Chimera Version
<th>Machine
<th>Processor

<th>Clock (GHz)
<th>Main Memory (Gb)
<th>OS Version
<th>Driver Version
<th>Host name

<tr><th><th><th>

<tr align="center">
<td>GeForce 9800 GTX
<td>512
<td>Windows
<td> 755 <td> 606  <td> 171 <td> 291 <td> 132

<td>60 <td>13.6 <td>13.5 <td> 36.8 <td> 28.8 <td> 37.4
<td>
<a href="../ticket/5680">5680</a>
(<a href="#note16">16</a>,<a href="#note17">17</a>)

<td>5/08
<td>2509
<td>Cyberpower
<td>Athlon 64 X2 5200+
<td>3.12
<td>3
<td>Windows XP Home SP2
<td>2.1.2
<td>from Sacha De Carlo

<tr align="center">
<td>Quadro FX 5600
<td>1536
<td>Mac
<td> 570 <td> 568  <td> 185 <td> 406 <td> 152

<td>168 <td>4.5 <td>4.6 <td>7.9 <td> 6.7 <td> 40.9
<td><a href="../ticket/5669">5669</a>
<td>5/08
<td>2509
<td>Mac Pro
<td>Dual Intel Xeon Quad-Core
<td>2.8
<td>10
<td>Mac OS 10.5.2

<td>2.0 nvidia 1.5.24
<td>strawson.cgl.ucsf.edu

<tr align="center">
<td>Radeon X1600
<td>256
<td>Mac
<td> 425 <td> 426  <td> 156 <td> 256 <td> 97
<td>140 <td>3.6 <td>3.6 <td>9.0 <td> 5.8 <td> 27.2

<td><a href="../ticket/5667">5667</a>
<td>5/08
<td>2509
<td>iMac
<td>Intel Core Duo
<td>2
<td>2
<td>Mac OS 10.4.11
<td>2.0 ATI 1.4.56
<td>oscar.cgl.ucsf.edu

<tr align="center">
<td>Quadro FX 3400
<td>
<td>Windows
<td> 402 <td> 592  <td> 151 <td> 203 <td> 121

<td>59.8 <td>11.0 <td>10.9 <td>23.2 <td> 23.9 <td> 23.0
<td><a href="../ticket/5676">5676</a>
<td>5/08
<td>2509
<td>HP xw6200
<td>
<td>
<td>2
<td>Windows XP Pro SP2

<td>2.0.3
<td>huck.cgl.ucsf.edu

<tr align="center">
<td>GeForce 6800 Ultra
<td>256
<td>Mac
<td> 347 <td> 300  <td> 155 <td> 289 <td> 77
<td>104 <td>4.7 <td>4.8 <td>11.2 <td> 7.9 <td> 18.1

<td><a href="../ticket/5675">5675</a>
<td>5/08
<td>2509
<td>Power Mac G5
<td>Dual PowerPC G5 (3.1)
<td>2.7
<td>2
<td>Mac OS 10.4.11
<td>1.5 ATI 1.4.18
<td>twain.cgl.ucsf.edu

<tr align="center">
<td>Radeon X1600
<td>256
<td>Mac
<td> 344 <td> 343  <td> 170 <td> 256 <td> 86

<td>165 <td>5.1 <td>5.7 <td>12.8 <td> 5.1 <td> 32.2
<td><a href="../ticket/5649">5649</a>
<td>5/08
<td>2509
<td>MacBook Pro
<td>Intel Core 2 Duo
<td>2.33
<td>2
<td>Mac OS 10.4.11

<td>2.0 ATI 1.4.56
<td>Tom Goddard's laptop

<tr align="center">
<td>Radeon 9800 Pro
<td>128
<td>Mac
<td> 323 <td> 323  <td> 146 <td> 256 <td> 109
<td>89.1 <td>3.4 <td>3.7 <td>9.9 <td> 6.2 <td> 14.1

<td><a href="../ticket/5668">5668</a>
<td>5/08
<td>2509
<td>Power Mac G5
<td>Dual PowerPC 970 (2.2)
<td>2
<td>2.5
<td>Mac OS 10.5.2
<td>2.0 ATI 1.5.24
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 980 XGL
<td>
<td>Windows
<td> 289 <td> 296  <td> 156 <td> 71 <td> 128

<td>60.0 <td>12.1 <td>12.0 <td>30.0 <td> 20.0 <td> 23.7
<td><a href="../ticket/5677">5677</a>
<td>5/08
<td>2509
<td>HP xw4100
<td>
<td>
<td>1
<td>Windows XP Pro SP2

<td>1.5.3
<td>exhibitus.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 900 XGL
<td>
<td>Linux
<td> 281 <td> 289  <td> 156 <td> 73 <td> 113
<td>85.2 <td>8.5 <td>8.4 <td>21.0 <td> 17.1 <td> 17.7

<td><a href="../ticket/5671">5671</a>
<td>5/08
<td>2509
<td>
<td>
<td>
<td>
<td>Linux 2.6.23.15-80.fc7 i686
<td>1.5.8 nvidia 96.43.01
<td>buckley.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 980 XGL
<td>
<td>Windows
<td> 240 <td> 262  <td> 158 <td> 152 <td> 114

<td>100 <td>14.1 <td>14.2 <td>31.4 <td> 22.2 <td> 25.0
<td><a href="../ticket/5670">5670</a>
<td>5/08
<td>2509
<td>HP xw4100
<td>
<td>
<td>2
<td>Windows XP Pro SP2

<td>1.5.3
<td>dali.cgl.ucsf.edu

<tr align="center">
<td>Radeon Mobility 9200
<td>32
<td>Mac
<td> 158 <td> 167  <td> 97 <td> 128 <td> 72
<td>44 <td>1.1 <td>1.1 <td>5.1 <td> 2.1 <td> 9.0

<td>
<td>6/08
<td>2509
<td>iBook
<td>G4
<td>1.2
<td>0.5
<td>Mac OS 10.4.11
<td>
<td>from Miguel Ortiz-Lombardia

<tr align="center">
<td>VPro B 128
<td>128
<td>IRIX
<td> 157 <td> 150  <td> 70 <td> 128 <td> 63

<td>20.1 <td>1.0 <td>1.0 <td>2.4 <td> 1.8 <td> 2.3
<td><a href="../ticket/5672">5672</a>
<td>5/08
<td>2509
<td>SGI Octane
<td>
<td>
<td>
<td>IRIX 6.5.24m IP30

<td>1.2
<td>euler.cgl.ucsf.edu

<tr align="center">
<td>GeForce 7300 GT
<td>256
<td>Mac
<td> 116 <td> 241  <td> 161 <td> 256 <td> 129
<td>60 <td>5.8 <td>5.7 <td>8.3 <td> 9.2 <td> 27.2

<td><a href="../ticket/5802">5802</a>
(<a href="#note18">18</a>)
<td>6/08
<td>2524 Aqua
<td>Mac Pro
<td>2 Dual-Core Intel Xeon
<td>2
<td>4
<td>Mac OS 10.5.3
<td>2.0 NVIDIA-1.5.28
<td>

</table>


<h3>
Scores for Chimera Versions &gt;= 1892
</h3>

<font color=red>
<p>

Some of these scores are incorrect high values.  Prior to 1.2509 Chimera version the benchmark code did not draw the models to the screen in 
in order to avoid syncronization to the display refresh rate (common with old
CRT displays).  Some
drivers were able to take advantage of this and shortcut the
full rendering giving incorrect timings.
</p>
</font>

<table border=1>

<tr>
<th>Graphics
<th>Processor
<th>Clock (MHz)
<th>OS Type
<th>Surface
<th>Mesh
<th>Contour
<th>Solid
<th>Recolor
<th>Notes

<th>Test Date
<th>Chimera Version
<th>Machine
<th>Graphics Memory (Mb)
<th>Main Memory (Mb)
<th>OS Version
<th>Driver Version
<th>Host name

<tr align="center">
<td>Quadro FX 3500
<td>2 dual-core Opteron 2220
<td>
<td>
<td> 842 <td> 846  <td> 153 <td> 477 <td> 153

<td align="left">
<td>8/07
<td>2422
<td>HP xw9400
<td>256
<td>8192
<td>
<td>100.14.11
<td>From Dave Konerding

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>46 <td>114 <td>9 <td>8 <td>(<a href="#note15">n/a</a>) <td> 25 <td> 21  
</table>

<tr align="center">
<td>Quadro FX 1500
<td>Intel Core 2 6600

<td>2400
<td>Linux
<td> 839 <td> 809  <td> 179 <td> 435 <td> 161
<td align="left">
<td>9/07
<td>2422
<td>
<td>256

<td>2048
<td>Fedora Core 6
<td>
<td>From Gabor Papai

<tr align="center">
<td>GeForce 8800 GT
<td>2 quad-core Intel Xeon
<td>2800
<td>Mac
<td> 766 <td> 754  <td> 184 <td> 406 <td> 239

<td align="left">
<td>3/08
<td>2500(?)
<td>Mac Pro
<td>512
<td>2048
<td>Mac OS 10.5.2(?)
<td>
<td>From Jeff Triffo

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>163 <td>170 <td>5 <td>5 <td>8 <td>7 <td> 42
</table>

<tr align="center">
<td>2 NVidia Quadro FX3400s w/SLI
<td>Dual Xeon
<td>3400
<td>Windows
<td> 709  <td> 709  <td> 156  <td> 372  <td> 138

<td align="left">PCI Express
<td>8/04
<td>1980
<td>3DBOXX 8206
<td>256 each
<td>2048
<td>XP Professional
<td>65.90 prerelease
<td>SIGGRAPH04 booth

<tr align="center">
<td>Radeon HD 2400 Pro
<td>Dual Intel E2160
<td>1800
<td>Windows
<td> 648  <td> 554  <td> 168  <td> 248  <td> 115

<td align="left">
<td>4/08
<td>2492
<td>Dell Inspiron 530s
<td>256
<td>3072
<td>Windows XP
<td>6.14.0010.6764
<td>From Sacha De Carlo.

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>93 <td>171 <td>8 <td>7 <td>(<a href="#note15">n/a</a>) <td>14 <td> 30
</table>

<tr align="center">
<td>Quadro FX 5600
<td>2 Quad-core Xeon
<td>2800

<td>Mac
<td> 633  <td> 627  <td> 184  <td> 443  <td> 239
<td align="left">Out of main memory at solid 444.
<td>3/08
<td>2496
<td>PowerMac
<td>1536
<td>2048

<td>Mac OS 10.5.2
<td>
<td>strawson.compbio.ucsf.edu

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops
<tr align=center><td>1F4H (33805)<td>161       <td>168       <td>5         <td>5 <td>(<a href="#note15">n/a</a>) <td>7 <td>40                     

</table>

<tr align="center">
<td>NVidia Quadro FX3400
<td>Dual Xeon
<td>3400
<td>Windows
<td> 631  <td> 672  <td> 152  <td> 300  <td> 136

<td align="left">PCI Express
<td>8/04
<td>1980
<td>3DBOXX 8208
<td>256
<td>2048
<td>XP Professional
<td>61.76
<td>SIGGRAPH04 booth

<tr align="center">
<td>Quadro FX 1400
<td>Pentium M
<td>2000
<td>Linux
<td> 601  <td> 682  <td> 156  <td> 310  <td> 150

<td align="left">
<td>10/06
<td>2199
<td>Dell Precision M70 laptop
<td>256
<td>1024
<td>Redhat Enterprise 4 WS
<td>1.0-8774
<td>CGL machine

<tr align="center">
<td>ATI FireGL X1 256p
<td>Dual Xeon
<td>2400
<td>Windows
<td> 592  <td> 594  <td> 142  <td> 298  <td> 85

<td align="left">2X multisampling
<td>4/04
<td>1917
<td>Compaq Evo W8000
<td>256
<td>1024
<td>XP Professional
<td>7.98.5
<td>whimsey.cgl.ucsf.edu

<tr align="center">
<td>ATI FireGL X1 256p
<td>Dual Xeon
<td>2400
<td>Windows
<td> 583  <td> 585  <td> 142  <td> 285  <td> 86

<td align="left">6X multisampling
<td>4/04
<td>1917
<td>Compaq Evo W8000
<td>256
<td>1024
<td>XP Professional
<td>7.98.5
<td>whimsey.cgl.ucsf.edu

<tr align="center">
<td>Quadro FX 1400
<td>Pentium M
<td>2000
<td>Windows
<td> 577  <td> 683  <td> 156  <td> 256  <td> 156

<td align="left">Solid size 300 - "not enough memory".
<td>10/06
<td>2255
<td>Dell Precision M70 laptop
<td>256
<td>1024
<td>Windows XP Pro
<td>84.30
<td>CGL machine

<tr align="center">
<td>Nvidia Quadro FX 4500
<td>Dual-core Xeon
<td>2660
<td>Mac
<td>  <td>  <td>  <td>  <td>

<td align="left">
<td>5/07
<td>2382
<td>Apple Mac Pro
<td>512
<td>2048
<td>Mac OS 10.4.9
<td>
<td>From Matthew Dougherty

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>154<td>166<td>8<td>8<td>(<a href="#note15">n/a</a>)<td>13<td>38              
</table>

<tr align="center">
<td>Radeon X1900
<td>2 Dual-core Xeon
<td>2660
<td>Mac

<td> 567  <td> 566  <td> 192  <td> 350+  <td> 134
<td align="left">Solid size &gt;350 - "not enough memory".
<td>1/07
<td>2304
<td>Apple Mac Pro
<td>512
<td>2048

<td>Mac OS 10.4.8
<td>
<td>From Jeff Speir

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops
<tr align=center><td>1F4H (33805)<td>150<td>167<td>6<td>6<td>4<td>10<td>38

</table>

<tr align="center">
<td>GeForce 7900GT
<td>Intel Core 2 Duo Extreme 6800
<td>2900
<td>Linux
<td> 549  <td> 472  <td> 194  <td> 489+  <td> 172

<td align="left">Out of memory on solid after size 489 at 12.5 fps.
<td>10/06
<td>2199
<td>Custom built.
<td>256
<td>4192
<td>SUSE 10.1
<td>x86_64 8774
<td>From David Konerding.

<tr align="center">
<td>Radeon X1600
<td>Intel Core Duo
<td>2000
<td>Mac
<td> 547  <td> 543  <td> 157  <td> 256  <td> 75

<td align="left">
<td>10/06
<td>2255
<td>iMac
<td>256
<td>1024
<td>Mac OS 10.4.8
<td>
<td>oscar.cgl.ucsf.edu

<tr align="center">
<td>Radeon X1600
<td>Intel Core 2 Duo
<td>2330
<td>Mac
<td> 541  <td> 541  <td> 170  <td> 272  <td> 93

<td align="left">
<td>3/08
<td>2496
<td>MacBook Pro
<td>256
<td>2048
<td>Mac OS 10.4.11
<td>
<td>feyerabend.cgl.ucsf.edu

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>159<td>174<td>4.5<td>5.7<td>13<td>5.2<td>34
</table>

<tr align="center">
<td>GeForce 7900GT
<td>Intel Core 2 Duo Extreme 6800
<td>2900
<td>Windows
<td> 507  <td> 489  <td> 211  <td> 349  <td> 172

<td align="left">
<td>10/06
<td>2199
<td>Custom built.
<td>256
<td>4192
<td>Win XP 32-bit
<td>91.47_forceware
<td>From David Konerding.

<tr align="center">
<td>Radeon HD 2600
<td>2 Quad-core Xeon
<td>2800
<td>Mac
<td> 468  <td> 466  <td> 182  <td> 325  <td> 217

<td align="left">
<td>2/08
<td>2470
<td>Apple Mac Pro
<td>256
<td>6144
<td>Mac OS 10.5
<td>
<td>From Jeff Triffo

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>152   <td>159   <td>6     <td>5     <td>(<a href="#note15">n/a</a>) <td>8     <td>40   
</table>

<tr align="center">
<td>3dLabs VP990 Pro
<td>P4
<td>3000
<td>Windows

<td> 431  <td> 428  <td> 156  <td> 250+  <td> 102
<td align="left">Driver set for max texture performance.
<td>4/04
<td>1917
<td>Dell OptiPlex GX270
<td>512
<td>1024
<td>XP Professional

<td>3.01-0739
<td>Greg's computer

<tr align="center">
<td>3dLabs VP990 Pro
<td>P4
<td>3000
<td>Windows
<td> 427  <td> 421  <td> 156  <td> 127  <td> 96

<td align="left">No multisampling.
<td>4/04
<td>1917
<td>Dell OptiPlex GX270
<td>512
<td>1024
<td>XP Professional
<td>3.01-0739
<td>Greg's computer

<tr align="center">
<td>ATI Mobility Radeon X1300
<td>Intel Core Duo T2300
<td>1700
<td>Windows
<td> 394  <td> 486  <td> 141  <td> 256  <td> 88

<td align="left">Laptop.
<td>4/06
<td>2199
<td>Dell Inspiron E1505
<td>128
<td>512
<td>Windows XP Media Edition
<td>
<td>From Daniil Zaitsev

<tr align="center">
<td>ATI Mobility Radeon 9700
<td>G4
<td>1670
<td>Mac
<td> 379  <td> 382  <td> 116  <td> 256  <td> 80

<td align="left">Laptop.
<td>3/05
<td>2065
<td>PowerBook
<td>128
<td>?
<td>10.3.8 ?
<td>
<td>Tom Ferrin's computer

<tr align="center">
<td>3dLabs VP990 Pro
<td>P4
<td>3000
<td>Windows
<td> 357  <td> 264  <td> 156  <td> 76  <td> 89

<td align="left">With 4x multisampling.
<td>4/04
<td>1917
<td>Dell OptiPlex GX270
<td>512
<td>1024
<td>XP Professional
<td>3.01-0739
<td>Greg's computer

<tr align="center">
<td>Radeon X1900
<td>2 Dual-core Xeon
<td>2660
<td>Windows
<td> 315  <td> 315  <td> 186  <td> 350+  <td> 129

<td align="left">Solid size &gt;350 - "not enough memory".
<td>1/07
<td>2304
<td>Apple Mac Pro
<td>512
<td>2048
<td>XP
<td>
<td>From Jeff Speir

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>92<td>169<td>10<td>10<td>25<td>18<td>42
</table>

<tr align="center">
<td>Radeon 9800 Pro
<td>Dual G5
<td>2000
<td>Mac
<td> 306  <td> 306  <td> 142  <td> 256  <td> 94

<td align="left">Chimera compiled on G4.
<td>4/04
<td>1917
<td>Power Mac
<td>128
<td>1536
<td>10.3.3
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 500 GoGL
<td>P4 Mobile
<td>2200
<td>Linux
<td> 296 <td> 170 <td> 138 <td> 217 <td> 108

<td align="left">Mesh (<a href="#note10">10</a>) and solid
(<a href="#note11">11</a>) unexpectedly slow.
<td>4/04
<td>1917
<td>Dell Precision M50 laptop
<td>64
<td>1024
<td>Redhat 8.0
<td>4363
<td>feyerabend.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 980 XGL
<td>P4
<td>3000

<td>Windows
<td> 287 <td> 287 <td> 158 <td> 256 <td> 120
<td align="left">
<td>4/04
<td>1917
<td>HP Workstation xw4100
<td>128
<td>1024

<td>XP Pro
<td>5664
<td>exhibitus.cgl.ucsf.edu

<tr align="center">
<td>Quadro NVS 285
<td>Dual P4
<td>3400
<td>Windows
<td> 261 <td> 116 <td> 144 <td> 110 <td> 95

<td align="left">
<td>9/07
<td>2422
<td>Dell Precision WorkStation 380
<td>64
<td>1024
<td>XP Pro
<td>7.1.8.4
<td>Dr. Dave's computer

<tr align="center">
<td>ATI FireGL 8800
<td>P4
<td>2400
<td>Linux
<td> 234 <td> 227 <td> 142 <td> 256 <td> 107

<td align="left">
<td>10/05
<td>2176
<td>HP XW5500 workstation
<td>128
<td>512
<td>Gentoo GNU/LINUX
<td>fglrx 8.14.30
<td>from Mingfeng Yang

<tr align="center">
<td>GeForce4 MX
<td>Athlon 2800
<td>2100
<td>Windows
<td> 163 <td> 153 <td> 131 <td> &gt;250 <td> 118

<td align="left">
<td>3/05
<td>2090
<td>Emachines T2825
<td>64
<td>512
<td>XP Home, SP2
<td>6.6.9.3
<td>TG's computer

<tr align="center">
<td>Radeon Mobility M7
<td>Pentium M
<td>1400
<td>Linux
<td> 147 <td> 127 <td> 133 <td> 128 <td> 68

<td align="left">
<td>10/05
<td>2176
<td>IBM T41
<td>32 (?)
<td>512
<td>Ubuntu GNU/LINUX 5.10
<td>Xorg-x11 6.8.2
<td>from Mingfeng Yang

<tr align="center">
<td>Intel 965GM
<td>Intel Core 2 duo
<td>1600
<td>Linux
<td> 130  <td> 129  <td> 158  <td> 128  <td> 129

<td align="left">
<td>8/07
<td>2422
<td>Thinkpad x61t
<td>256
<td>1024
<td>2.6.22
<td>xf86-video-intel 2.1.1
<td>From Damir Perisa

<tr><td><td><td><td>
<td colspan=14>
<table border=1 rules=cols cellpadding=5>
<tr><td>Molecule (Atoms)<td>Null<td>Wire<td>Stick<td>BStick<td>Ribbon<td>Sphere<td>Ops

<tr align=center><td>1F4H (33805)<td>97    <td>53    <td>1     <td>1     <td>(<a href="#note15">n/a</a>) <td>2     <td>28   
</table>

<tr align="center">
<td>Radeon Mobility 7500
<td>P4
<td>1600
<td>Windows

<td> 119 <td> 122 <td> 110 <td> 99 <td> 63
<td align="left">
<td>10/05
<td>2172
<td>Compaq Evo N800V
<td>32
<td>512
<td>XP

<td>
<td>from Boaz Shaanan

<tr align="center">
<td>Intel GMA 950
<td>Intel Core 2 Duo
<td>1830
<td>Mac
<td> 112 <td> 121 <td> 151 <td> 211 <td> 98

<td align="left">No hardware T&L.
<td>10/06
<td>2255 ?
<td>iMac.  Same GPU in Mac Mini.
<td>64
<td>512
<td>10.4.8
<td>
<td>from Greg Couch

</table>

<h3>
Scores for Chimera Versions 1711 to 1891
</h3>

<p>

Chimera versions 1892 and later use two light sources instead of one,
which reduces the surface and mesh rendering scores.
</p>

<table border=1>
<tr>
<th>Graphics
<th>Processor
<th>Clock (MHz)
<th>OS Type
<th>Surface
<th>Mesh
<th>Contour
<th>Solid
<th>Recolor
<th>Notes
<th>Test Date
<th>Chimera Version

<th>Machine
<th>Graphics Memory (Mb)
<th>Main Memory (Mb)
<th>OS Version
<th>Driver Version
<th>Host name

<tr align="center">
<td>Quadro4 900 XGL
<td>Xeon
<td>2800
<td>Windows XP
<td> 384 <td> 375 <td> 150 <td> 256 <td> 123

<td nowrap align="left">
<td>6/03
<td>1741
<td>hp workstation xw6000
<td>128
<td>1024
<td>XP Professional
<td>4.0.7.2
<td>cgl-test9.cgl.ucsf.edu

<tr align="center">
<td>Radeon 9800 Pro
<td>Dual G5
<td>2000
<td>Mac
<td> 381  <td> 378  <td> 138  <td> 256  <td> 93

<td align="left">Chimera compiled on G4.
<td>6/03
<td>1747
<td>Power Mac
<td>128
<td>2048
<td>10.2.7
<td>
<td>WWDC demo machine

<tr align="center">
<td>Radeon 9800 Pro
<td>Dual G5
<td>2000
<td>Mac
<td> 374  <td> 374  <td> 142  <td> 256  <td> 103

<td align="left">Chimera compiled on G4.
<td>11/03
<td>1872
<td>Power Mac
<td>128
<td>1536
<td>10.3
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>GeForce 4 Ti 4600
<td>Dual G4
<td>1000
<td>Mac
<td> 349 <td> 347 <td> 102 <td> 256  <td> 67

<td align="left">
<td>6/03
<td>1747
<td>Power Mac
<td>64
<td>1280
<td>10.2.6
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Quadro4 500 GoGL
<td>P4 Mobile
<td>2200
<td>Linux
<td> 310 <td> 179 <td> 134 <td> 196 <td> 109

<td align="left">Mesh (<a href="#note10">10</a>) and solid
(<a href="#note11">11</a>) unexpectedly slow.
<td>6/03
<td>1747
<td>Dell Precision M50 laptop
<td>64
<td>1024
<td>Redhat 8.0
<td>4363
<td>feyerabend.cgl.ucsf.edu

</table>

<h3>
Scores for Chimera Versions &lt; 1711

</h3>

<p>
The scores in the following table are for Chimera versions prior to 1711.
Optimizations for surface and mesh rendering
(<a href="#note12">12</a>)
were added in Chimera version 1711 and contouring
(<a href="#note13">13</a>)
in Chimera 1716 that greatly increased scores.
</p>

<table border=1>
<tr>
<th>Graphics
<th>Processor
<th>Clock (MHz)
<th>OS Type
<th>Surface

<th>Mesh
<th>Contour
<th>Solid
<th>Recolor
<th>Notes
<th>Test Date
<th>Chimera Version
<th>Machine
<th>Graphics Memory (Mb)
<th>Main Memory (Mb)
<th>OS Version
<th>Driver Version
<th>Host name

<tr align="center">
<td>GeForce 4 Ti 4600
<td>Dual G4
<td>1000

<td>Mac
<td> 290  <td> 290  <td> 70  <td> 256  <td> 67
<td align="left">Using prerelease Mac OS 10.2.5 build 6L14.
<td>3/03
<td>1700
<td>Power Mac
<td>64
<td>1280

<td>10.2.5
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Radeon 9000
<td>Dual G4
<td>1000
<td>Mac
<td> 249  <td> 254  <td> 70  <td> 228  <td> 75

<td align="left">Using prerelease Mac OS 10.2.5 build 6L14.
<td>3/03
<td>1700
<td>Power Mac
<td>64
<td>1280
<td>10.2.5
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>NVidia Quadro4 900 XGL
<td>Dual Xeon
<td>2400
<td>Windows XP
<td> 223  <td> 223  <td> 104  <td> 256  <td> 118

<td nowrap align="left">
Solid of size 256 gives<br>
51.4 frames per second,<br>
size 257 gives 0.9 fps.
<a href="#note1">(1)</a><br>
<td>10/02
<td>1592
<td>Compaq Workstation W8000
<td>128
<td>1024
<td>XP Professional
<td>3.0.8.7
<td>whimsey.cgl.ucsf.edu

<tr align="center">

<td>GeForce 4 Ti 4400
<td>P4
<td>1800
<td>Linux
<td> 212  <td> 201  <td> 98  <td> 256  <td> 100
<td nowrap align="left">
Solid of size 256 gives<br>

20 frames per second,<br>
size 257 gives 3 fps.
<a href="#note1">(1)</a><br>
Also see
<a href="#note2">(2)</a>,
<a href="#note3">(3)</a>.
<td>8/02
<td>1522
<td>
<td>128
<td>1024
<td>Debian 3.0
<td>NVIDIA 28.80

<td>tamarin.ucsf.edu

<tr align="center">
<td>FireGL 2
<td>P4
<td>1400
<td>Windows 2000
<td> 209 <td> 208 <td> 47 <td> 61 <td> 92
<td align="left">Slow contouring in Chimera 1602

<a href="#note9">(9)</a>.
<td>12/02
<td>1602
<td>Compaq Deskpro
<td>64
<td>768
<td>2000 Professional
<td>5.12.2195.2082
<td>arnauld.cgl.ucsf.edu

<tr align="center">
<td>FireGL 2
<td>P4
<td>1400
<td>Windows 2000
<td> 208 <td> 207 <td> 76 <td> 63 <td> 79

<td>
<td>7/02
<td>1516
<td>Compaq Deskpro
<td>64
<td>768
<td>2000 Professional
<td>5.12.2195.2082
<td>arnauld.cgl.ucsf.edu

<tr align="center">
<td>FireGL 2
<td>P4
<td>1400
<td>Linux
<td> 194  <td> 194  <td> 82  <td> 62  <td> 78

<td>
<td>7/02
<td>1516
<td>Compaq Deskpro
<td>64
<td>768
<td>Redhat 7.2
<td>firegl23 1.9.7
<td>tolkien.cgl.ucsf.edu

<tr align="center">
<td>FireGL 2
<td>P4
<td>2000
<td>Linux
<td> 192  <td> 194  <td> 100  <td> 60  <td> 92

<td align="left">Scores reported by Dave Konerding.
<td>9/02
<td>1516
<td>White box PC
<td>64
<td>1024
<td>Redhat 7.3
<td>FireGL2 1.9.20
<td>

<tr align="center">
<td>Quadro4 500 GoGL
<td>P4 Mobile
<td>2200
<td>Windows XP
<td> 179 <td> 179  <td> 71 <td> 256 <td> 115

<td align="left">
Crashed on solid and recolor
unless run separately.
Also see
<a href="#note1">(1)</a>,
<a href="#note9">(9)</a>.
<td>12/02
<td>1602
<td>Dell Precision M50 laptop
<td>64
<td>1024
<td>XP Professional
<td>6.13.10.2835
<td>

<tr align="center">
<td>Quadro4 500 GoGL

<td>P4 Mobile
<td>2200
<td>Linux
<td> 176 <td> 176  <td> 94 <td> 256 <td> 106
<td align="left">
Solid of size 256 gives<br>
18 frames per second.

<a href="#note1">(1)</a><br>
<td>12/02
<td>1602
<td>Dell Precision M50 laptop
<td>64
<td>1024
<td>Redhat 8.0
<td>4191
<td>

<tr align="center">
<td>GeForce 2 GTS/Pro
<td>Athlon
<td>1300
<td>Windows XP
<td> 165  <td> 169  <td> 75  <td> 256  <td> 96

<td align="left">Solid of size 256 gives 18 frames per second.
<a href="#note1">(1)</a>
Also see
<a href="#note2">(2)</a>.
<td>7/02
<td>1516
<td>
<td>64
<td>512
<td>XP
<td>2.8.3.2
<td>chagall.nmr.ucsf.edu

<tr align="center">
<td>VPro 12
<td>R12000

<td>300
<td>IRIX
<td> 157  <td> 157  <td> 57  <td> 146  <td> 69
<td>
<td>7/02
<td>1516
<td>SGI Octane
<td>128

<td>640
<td>6.5.14f
<td>
<td>larson.cgl.ucsf.edu

<tr align="center">
<td>Quadro2 Go (Dell Mobile)
<td>P3 Mobile
<td>1200
<td>Windows 2000
<td> 140  <td> 140  <td> 81  <td> 128  <td> 66

<td align="left">Crashes on solid test.
<a href="#note5">(5)</a>
<td>10/02
<td>1587
<td>Dell Precision M40 laptop
<td>
<td>512
<td>2000
<td>1.4.2.7
<td>

<tr align="center">
<td>ELSA Gloria II Quadro
<td>Dual P3?
<td>866
<td>Windows NT

<td> 140  <td> 139  <td> 66  <td> 57  <td> ?
<td align="left">Erratic solid display frame rates.
Also see
<a href="#note2">(2)</a>.
<td>7/02
<td>1516
<td>Compaq AP550
<td>64

<td>256
<td>NT
<td>4.00.1381.2832
<td>edman.cgl.ucsf.edu

<tr align="center">
<td>ATI Radeon 9000
<td>Dual G4
<td>1000
<td>Mac
<td> 119  <td> 119  <td> 66  <td> 240  <td> 75

<td align="left">Using prerelease Mac OS X Aqua
Chimera.
<a href="#note8">(8)</a>.
<td>12/02
<td>1619
<td>Power Mac
<td>64
<td>1280
<td>10.2.2
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>ATI Radeon 9000
<td>Dual G4
<td>1000
<td>Mac
<td> 118  <td> 118  <td> 71  <td> 239  <td> 57

<td align="left">Using XFree86 4.2.99.1
<a href="#note7">(7)</a>.
<td>10/02
<td>1574
<td>Power Mac
<td>64
<td>1280
<td>10.2.1
<td>ATI Radeon 8500 OpenGL Engine
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Radeon 9000 Mobility
<td>G4
<td>1000
<td>Mac
<td> 98  <td> 98  <td> 53  <td> 227  <td> 64

<td align="left">Using Apple's X11 beta.
Results from Rey Banatao.
<td>2/03
<td>1602
<td>PowerBook
<td>64
<td>1024
<td>10.2.3
<td>
<td>

<tr align="center">
<td>Infinite Reality 2
<td>Dual R12000
<td>400
<td>IRIX
<td> 94  <td> 87  <td> 65  <td> 48  <td> 68

<td>
<td>7/02
<td>1516
<td>SGI Onyx
<td>64
<td>1024
<td>6.5.14f
<td>
<td>spinoza.cgl.ucsf.edu

<tr align="center">
<td>GeForce4 420 Go (<a href="#note14">14</a>)
<td>G4
<td>867
<td>Mac
<td> 78  <td> 84  <td> 63  <td> 143  <td> 64

<td align="left">Apple X11 beta 3
<td>7/03
<td>1700
<td>12" PowerBook
<td>32
<td>640
<td>10.2.6
<td>
<td>Patsy Babbit's machine

<tr align="center">
<td>GeForce 2 MX
<td>G4
<td>733
<td>Mac
<td> 70  <td> 76  <td> 60  <td> 48  <td> 40

<td align="left">Using XFree86 4.2.99.1
<a href="#note7">(7)</a>.  Display 1600x1024.
<td>10/02
<td>1574
<td>Power Mac
<td>32
<td>768
<td>10.2.1
<td>
<td>essex.cgl.ucsf.edu

<tr align="center">
<td>Cobalt
<td>Dual P2
<td>600
<td>Windows NT
<td> 70  <td> 70  <td> 57  <td> 85  <td> 48

<td>
<td>7/02
<td>1516
<td>SGI 320
<td>
<td>256
<td>NT
<td>5.2.3.000, 5.2.3
<td>ryle.cgl.ucsf.edu

<tr align="center">
<td>PowerStorm 350
<td>Dual Alpha EV6.7
<td>667
<td>Tru64
<td> 66  <td> 73  <td> 73  <td> 50  <td> 71

<td align="left">Solid rendering shows white cube when using glFog().
Also see
<a href="#note2">(2)</a>.
<td>7/02
<td>1516
<td>Compaq DS-20
<td>64
<td>2096
<td>5.1
<td>
<td>retief.cgl.ucsf.edu

<tr align="center">
<td>ATI Radeon 7500 Mobility
<td>G4
<td>800
<td>Mac
<td> 63  <td> 63  <td> 65  <td> 128  <td> 30

<td align="left">
Using Xtools from Tenon <a href="#note6">(6)</a>.
Results from Mike Lowenstein.
<td>8/19
<td>1538
<td>PowerBook
<td>32
<td>1024
<td>10.2
<td>
<td>

<tr align="center">
<td>Voodoo 5 5500
<td>P3
<td>800

<td>Linux
<td> 54  <td> 44  <td> 67  <td> 106  <td> 54
<td align="left">Solid rendering shows texture junk on small 137 x 138 x 15 data when rotated
<td>7/02
<td>1516
<td>
<td>64
<td>~380

<td>Debian 3.0
<td>
<td>drill.ucsf.edu

<tr align="center">
<td>Mesa 3.4.2 / Radeon 9000
<td>Dual G4
<td>1000
<td>Mac
<td> 45  <td> 45  <td> 70  <td> 1  <td> 2

<td align="left">Using XDarwin 4.2.0.1.
No hardware acceleration
<a href="#note4">(4)</a>.
<td>9/02
<td>1574
<td>Power Mac
<td>64
<td>1280
<td>10.2.1
<td>
<td>mach.cgl.ucsf.edu

<tr align="center">
<td>Solid Impact
<td>R10000
<td>195
<td>IRIX
<td> 43  <td> 41  <td> 49  <td> 2  <td> 14

<td>
<td>7/02
<td>1516
<td>SGI Octane
<td>
<td>256
<td>6.5.15m
<td>
<td>monet.nmr.ucsf.edu

<tr align="center">
<td>CRM
<td>R5000
<td>300
<td>IRIX
<td> 32  <td> 34  <td> 45  <td> 28  <td> 45

<td>
<td>7/02
<td>1516
<td>SGI O2
<td>
<td>128
<td>6.5.6m
<td>
<td>utopia3.ucsf.edu

<tr align="center">
<td>Mesa 3.4.2 / Matrox G400
<td>P3
<td>800
<td>Linux
<td> 30  <td> 35  <td> 69  <td> 2  <td> 15

<td align="left">No hardware acceleration
<a href="#note4">(4)</a>.
<td>7/02
<td>1516
<td>
<td>
<td>~380
<td>Debian 3.0
<td>
<td>langur.ucsf.edu

<tr align="center">
<td>4D51T
<td>Alpha
<td>180
<td>Tru64
<td> 19 <td> 18 <td> 29 <td> 45 <td> 35

<td align="left">Solid and recolor benchmarks sometimes cause system hang.
<td>7/02
<td>1516
<td>Compaq AlphaStation 200 4/233
<td>
<td>128
<td>5.1
<td>
<td>feyerabend.cgl.ucsf.edu

<tr align="center">
<td>GR2
<td>R5000
<td>180
<td>IRIX
<td> 13 <td> 14 <td> 37 <td> 1 <td> 9

<td align="left">
<td>7/02
<td>1516
<td>SGI INDY
<td>
<td>64
<td>6.5.15m
<td>
<td>lautrec.nmr.ucsf.edu

</table>

<h2>
Notes
</h2>

<p>
<a name="note1">

<b>(1)</b>
</a>
Textures must have power of 2 sizes.  Going from size 256 to 257 makes
the 2 dimensional textures jump from size 256 by 256 to 512 by 512.  The
bigger textures may not fit in texture memory reducing performance
significantly.  A better implementation could render a data set with
edge size 257 using several smaller blocks to avoid this big jump in
performance.
</p>

<p>
<a name="note2">
<b>(2)</b>
</a>
These systems have a bug in the OpenGL implementation where glPixelMap()
does not work when loading a texture.  A fix to work around the problem
was put into Chimera 1522 but the speed of changing thresholds in colormap
solid rendering is compromised.
The colormap solid rendering is not tested by the benchmarks but is the
default solid rendering mode in the Chimera volume viewer extension.
No data is shown in colormap solid mode on these (pre 1522) systems,
but an outline box can be seen and rotated normally.
The colormap mode can be turned off in the volume viewer options panel.
The direct RGBA texture filling mode is then used which is slower when
changing thresholds.
</p>

<p>
<a name="note3">
<b>(3)</b>

</a>
Turning on the smooth mesh lines option makes the mesh lines black.
This is probably a graphics driver bug.
</p>

<p>
<a name="note4">
<b>(4)</b>
</a>
Mesa is a software implementation of OpenGL.
It does not use hardware acceleration provided by the graphics card.
The scores for systems using Mesa are probably determined by computer
CPU speed rather than graphics card performance.
</p>

<p>
<a name="note5">
<b>(5)</b>
</a>

Crashed on solid test after calculating frame rate for several data sizes.
Crash may occur only for volume size greater than 128.  Test solid model
displays  correctly.
</p>

<p>
<a name="note6">
<b>(6)</b>
</a>
Xtools X server from Tenon was reported to behave badly running Chimera.
Menus sometimes appear behind windows.  Volume viewer dialog window
sometimes resizes to zero size and must be resized by hand.  Many OpenGL
error messages written to shell by Chimera.
</p>

<p>
<a name="note7">
<b>(7)</b>
</a>
Compiled XFree86 4.2.99.1 from Oct 21, 2002 source code.  This
version of XFree86 has hardware OpenGL acceleration for Mac OS X.
X server crashed when started with OroborOSX 0.8b2 built for
XFree86 4.2.0.1 when I moved the Chimera window.  Worked when X server
started with startx command.

</p>

<p>
<a name="note8">
<b>(8)</b>
</a>
Prerelease Aqua Chimera causes Mac OS 10.2.2 system crash when
displaying side view window and showing 1gcn when switching to sphere
representation.  This crash also occurs on the X windows version of
Chimera for the Mac.  It is probably a Radeon 9000 graphics driver bug.
This does not effect benchmarks.
</p>

<p>
<a name="note9">
<b>(9)</b>
</a>
Chimera 1602 for Windows appears to have had the contouring code
compiled without optimization making it 5 times slower to calculate
isosurfaces than version 1516.  This effects the speed of the volume
viewer extension.
</p>

<p>
<a name="note10">
<b>(10)</b>
</a>
Mesh rendering did not improve on Quadro4 500 GoGL when limiting
display list size.  Factor of 3 speed ups were observed on other
Quadro4 and GeForce4 cards.  The NVidia driver apparently behaves
differently for this card.
</p>

<p>
<a name="note11">
<b>(11)</b>
</a>
Solid rendering with the NVidia 4363 driver appears to be slower than
with the version 4191 driver on Quadro4 500 GoGL card.
</p>

<p>
<a name="note12">
<b>(12)</b>
</a>
Surface and mesh rendering is about a factor of 3 faster on GeForce4
cards with early 2003 drivers (4191 linux) if display lists are
limited to less than about 64K triangles or lines.  I couldn't find
anything on the web describing this.  I changed the Chimera isosurface
and mesh rendering to use multiple display lists with limited size.
</p>

<p>
<a name="note13">
<b>(13)</b>
</a>
The contouring benchmark was bottle-necked by slow STL::map<> behaviour.
This was especially severe on Windows.  The map was used to avoid making
multiple copies of triangle vertices.  A more efficient array based method
increased the calculation speed by a factor of 3 for a 100^3 grid on most
platforms, and a factor of 10 on Windows.
Larger grids benefit less from this optimization.
</p>

<p>
<a name="note14">
<b>(14)</b>
</a>
Apple web pages say the 12" PowerBook has GeForce4 420 Go graphics.
The Apple system profiler program identifies the graphics as GeForce4 MX.
</p>

<p>
<a name="note15">
<b>(15) N/A ribbon scores</b>
</a>
The ribbon score on the molecule benchmark was not valid for Chimera versions
1.2382 through 1.2501 because a bug caused the ribbon not to be displayed
during the timing.  Only atoms as dots were shown.
</p>

<a name="note16">
<p>
<b>(16)</b>
Peak frame rate is limited in molecule benchmarks to 60 frames per second
because of synchronization to the display refresh rate.
</p>

<a name="note17">
<p>
<b>(17)</b>
CPU and graphics both overclocked.
</p>

<a name="note18">
<p>
<b>(18)</b>
Low surface/mesh scores probably because the machine was being used for
other tasks while benchmarks were running.  Detailed scores showed larger
test sizes producing faster frame rates.
</p>

}}}