Speedmark is Macworld Lab’s standard test tool for benchmarking new and upgraded systems. It uses real-world applications and everyday tasks. It is a general-purpose suite that includes tasks that everyone from a power user to a Mac rookie performs every day.
Each time major update to the Mac operating system introduces changes that require us to update our testing. Along with updating the OS version, we also update the individual applications to their latest shipping version and tweak and add tests to help create a better snapshot of a Mac’s overall performance.
Here’s a few of the changes we’ve made this time around:
• We added Create Archive and Unarchive tasks to the suite. We take a 1GB folder and use OSX’s built-in zip utility to create a zip file. We then restart the system and unzip the file we just created. This test tasks both the drives and processors and replaces our previous StuffIt test.
• We removed our CD Ripping to AAC via iTunes test due to inconsistencies we found when running the test, especially with the slot-loading drives found in laptops. This test also would penalize systems with DVD burners, which often had slower read speeds than CD-RW drives.
• We upgraded from Unreal Tournament 2003 to Unreal Tournament 2004 and now use the SantaDuck standard bench utility to test each system’s graphics card and each processors’ ability to run challenging 3-D games.
With all of these changes, it’s impossible to compare results from previous versions of Speedmark with results from Speedmark 4. So we’ve been busy here in the lab not only with tests of the latest Power Mac G5s, but with retests for a bunch of systems incorporating the new version of Speedmark.
Here’s a large benchmark chart that we hope will give you a good idea of the performance of currently available Macs under Speedmark 4.
Power Mac G5 Benchmarks
| Speedmark 4 | Cinema4D 9.1 | iMovie HD Render | iTunes 4.7 MP3 Encode | Photoshop CS2 Suite | Unreal Tournament 2004 | Compressor MPEG2 Encode | |
|---|---|---|---|---|---|---|---|
| Power Mac G5 Dual 2.7GHz | 250 | 1:02 | 0:25 | 0:52 | 0:53 | 49 | 3:52 |
| Power Mac G5 Dual 2.3GHz | 226 | 1:13 | 0:35 | 0:58 | 0:59 | 40 | 4:24 |
| Power Mac G5 Dual 2.0GHz | 205 | 1:24 | 0:29 | 1:08 | 1:06 | 38 | 4:46 |
| Power Mac G5 Dual 2.5GHz | 225 | 1:08 | 0:36 | 0:54 | 0:56 | 52 | 4:19 |
| Power Mac G5 Dual 2.0GHz (old) | 196 | 1:25 | 0:38 | 1:07 | 1:23 | 39 | 4:59 |
| Power Mac G5 1.8GHz | 160 | 3:06 | 0:40 | 1:49 | 1:33 | 27 | 10:04 |
| iMac G5 1.8GHz | 155 | 3:07 | 0:38 | 1:49 | 1:35 | 23 | 10:12 |
| iMac G5 1.6GHz | 141 | 3:28 | 1:00 | 2:03 | 1:45 | 23 | 11:16 |
| eMac 1.25GHz | 117 | 5:38 | 0:55 | 2:56 | 2:02 | 13 | 12:56 |
| Mac mini 1.42GHz | 115 | 4:54 | 1:01 | 2:33 | 1:47 | 12 | 14:16 |
| Mac mini 1.25GHz | 100 | 5:33 | 1:22 | 2:50 | 2:03 | 11 | 15:38 |
| PowerBook 1.67GHz | 135 | 4:27 | 0:52 | 2:20 | 1:35 | 20 | 12:46 |
| PowerBook 1.5GHz | 123 | 5:00 | 0:55 | 2:34 | 1:50 | 19 | 14:05 |
| iBook 1.33GHz | 100 | 5:16 | 1:08 | 2:44 | 1:59 | 11 | 14:28 |
| iBook 1.2GHz | 90 | 5:53 | 1:12 | 3:03 | 2:11 | 11 | 15:09 |
Best results in bold. Systems in italics are no longer in the product line. Included only for reference.
Speedmark 4 scores are relative to those of a 1.25GHz Mac mini which is assigned a score of 100. Photoshop, Cinema 4D, iMovie, and iTunes scores are in minutes:seconds. All systems were running Mac OS X 10.4, had 512MB RAM, with Processor Performance set to Highest in the Energy Saver pane in System Preferences. We converted 45 minutes of AAC audio files to MP3 using iTune’s High Quality settings. We used Unreal Tournament’s Antalus Botmatch Average Frames Per Second score tested at a resolution of 1,024-by-768 pixels at highest settings. The Photoshop Suite test is a set of 14 scripted tasks using a 50MB file. Photoshop’s memory was set to 70 percent and History was set to Minimum. Go here for more information on Speedmark 4.—Macworld Lab testing by James Galbraith and Jerry Jung.
For more on Speedmark—including a complete task list—visit our Speedmark page. And while Speedmark gives an excellent indication of real-world performance, it isn’t our only benchmark. We will continue to develop and include tests in a variety of applications appropriate for the hardware we review. If you have suggestions for tests that you‘d like to see run by Macworld Lab, let us know.