Now that the week-long frenzy of
Macworld Expo is in the rear-view mirror, Macworld Lab can turn its attention back to what we were working on prior to the annual Mac trade show—Speedmark results for the
new Mac Pro systems announced the first week of January.
The Friday before Expo began, we posted
our test results for Apple’s “recommended” configuration—an eight-core 2.8GHz system. The $2,799 system ships with 2GB of RAM, a 320GB hard drive and an ATI 2600 graphics card with 256MB of video memory. We found that the 2.8GHz eight-core model easily out-paced its four-core, 2.6GHz predecessor while coming close to matching the speed of a previous-generation eight-core 3GHz Mac Pro.
We’ve since received three different build-to-order systems: a 3.0GHz, eight-core Mac Pro ($3,599) as well as a 2.8GHz four-core system ($2,299). We also received a 3.2GHz eight-core system ($4,399), but had to return it as the system would not boot once we got it out of the packaging.
In the future, we hope to provide you with results using some of the optional graphics cards and RAID configurations, but at this time those options would have held up the delivery of our systems by two to three weeks, so we decided to get in what we could and work on getting the upgrades as soon as they become available.
In the meantime, let’s look at the results for the systems we have on hand.
More Mac Pro Benchmarks
| ||Speed- mark 5 ||Adobe PS CS3 ||Cinema 4D XL 10.5 ||Comp. 3 ||iMovie HD ||iTunes 7.5 ||Unreal Tourney 2004 ||Quake 4 ||Finder ||H’Brake ||Pro App Multi 2GB RAM ||Pro App Multi 4GB RAM |
| ||OVERALL SCORE ||SUITE ||RENDER ||MPEG2 ENCODE ||AGED EFFECT ||MP3 ENCODE || FRAME RATE || FRAME RATE ||ZIP ARCHIVE ||H.264 ENCODE ||SUITE ||SUITE |
|Mac Pro Xeon/2.8GHz (4 cores)* ||294 ||0:50 ||0:26 ||1:01 ||0:31 ||0:46 ||102.7 ||33.4 ||3:53 ||1:30 ||1:23 ||1:19 |
|Mac Pro Xeon/3GHz (8 cores)* ||327 ||0:40 ||0:15 ||0:52 ||0:30 ||0:44 ||106.4 ||32.9 ||3:43 ||1:02 ||1:06 ||1:03 |
|Mac Pro Xeon/2.8GHz (8 cores) ||314 ||0:49 ||0:16 ||0:46 ||0:30 ||0:47 ||100.8 ||33.3 ||3:51 ||0:54 ||1:17 ||1:11 |
| Mac Pro Xeon/2.66GHz (4 cores) ||258 ||0:51 ||0:28 ||1:20 ||0:39 ||0:52 ||87.8 ||51.4 ||5:15 ||1:55 ||1:39 ||1:27 |
| Mac Pro Xeon/3GHz (8 cores) (2007)* ||317 ||0:44 ||0:15 ||1:02 ||0:34 ||0:51 ||95.4 ||51.4 ||3:41 ||0:57 ||1:20 ||1:12 |
| Power Mac G5/2.5GHz (4 cores) ||225 ||1:12 ||0:32 ||1:28 ||0:37 ||0:48 ||32.6 ||37.7 ||4:47 ||2:25 ||2:28 ||2:23 |
| ||>Better ||<Better ||<Better ||<Better ||<Better ||<Better ||>Better ||>Better ||<Better ||<Better ||<Better ||<Better |
Best results in bold. Discontinued reference systems in italics. * Denotes build-to-order configuration.
Speedmark 5 scores are relative to those of a 1.5GHz Core Solo Mac mini, which is assigned a score of 100. Adobe Photoshop, Cinema 4D XL, iMovie, iTunes, and Finder scores are in minutes:seconds. All systems were running Mac OS X 10.5.1 with 2GB of RAM. 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. We recorded how long it took to render a scene in Cinema 4D XL. We used Compressor to encode a 6 minute, 26 second DV file using the DVD: Fastest Encode 120 minutes – 4:3 setting. In iMovie, we applied the Aged Film effect from the Video FX. menu to a one minute movie. We converted 45 minutes of AAC audio files to MP3 using iTunes’ High Quality setting. We used Unreal Tournament 2004’s Antalus Botmatch average-frames-per-second score; we tested at a resolution of 1,024 by 768 pixels at the Maximum setting with both audio and graphics enabled. We created a Zip archive in the Finder from a 2GB folder. For the Professional Application Multitasking suite, we recorded how long it took Photoshop to run our standard test suite while a longer Cinema 4D task and our Compressor encode test ran in the background.—MACWORLD LAB TESTING BY JAMES GALBRAITH, JERRY JUNG, AND BRIAN CHEN
The new eight-core, 3GHz system was about four percent faster in Speedmark 5, our overall system performance benchmark, than the eight-core 2.8GHz recommended configuration. The new 3GHz model was 13 percent faster in Compressor and a whopping 18 percent faster in Photoshop than the base model.
This new system is not the first eight-core, 3GHz Mac Pro offered by Apple; the company offered a similarly-rated model as a $3,997 build-to-order option, which we
reviewed last July. When comparing that older version with the new 3GHz eight-core system, we see the new system’s internal improvements at work. Overall, the new 3GHz model was a little more than 3 percent faster than its older 3GHz counterpart in Speedmark. Some individual tests showed more improvement, like a 9-percent performance gain in Photoshop and a 16-percent improvement in Compressor. Cinema 4D times were identical between the two 3GHz machines.
The other new system we received was a somewhat stripped-down version of the recommended configuration—it’s identical, with the exception that one of the four-core 2.8GHz Xeon processors has been removed. So how does a system fare with the same speed, but half the number of processor cores? It depends on the test.
As far as overall system performance goes, Speedmark 5 — which includes a mix of tests, some of which stress multiple processor cores and others of which do not — shows the new four-core 2.8GHz system to be about 6 percent slower than the eight-core 2.8GHz Mac Pro. In many of the individual tests, the systems finished within a second or two of each other, though Cinema 4D was about 63 percent slower on the four-core and Compressor was about 33 percent slower.
Though the 2.8GHz four-core Mac Pro is a build-to-order entry-level system, it features pretty impressive specifications when compared to the line’s previous recommended configuration—the four-core 2.66GHz Mac Pro, which debuted at a price of $2,499. When comparing these two systems’ performance, we see that the new system is 14 percent faster in Speedmark than the four-core 2.66GHz model, 24 percent faster in Compressor and 22 percent faster at HandBrake encoding. Photoshop and Cinema 4D scores were very similar between the two.
In our last Mac Pro Speedmark write-up, we promised to include more graphics tests and try running our professional application multitasking suite with more RAM. To that end, we ran Quake 4 and found the new Mac Pros, with their ATI HD 2600 XT cards, lagging behind the Nvidia GeForce 7300 found in the last generation of Mac Pros. This was not the case in Unreal Tournament, however, as the new ATI cards were able to squeeze about 30 more frames per second out of UT than the older Nvidia cards.
As far as the multitasking test, adding more RAM did help, allowing the 2.8GHz, eight-core system outfitted with 4GB of RAM to finish the test 16 percent faster than with the stock 2GB of memory. The performance benefit was less spectacular with the new 3GHz Mac Pro, getting only about a five-percent boost from the extra RAM, though the new 3GHz system’s multitasking test times were 13 percent faster than last year’s 3GHz Mac Pro model. In fact, we had to tweak the Cinema 4D test we use as part of our multitasking suite because the eight-core models simply finished it too quickly. All of the systems in the above table use the new, longer test settings.
At the request of many readers, we’ve added a Power Mac G5 in the chart for reference. As you can see, the new standard configuration Mac Pro (2.8GHz 8-core) was 40 percent faster than the 2.5GHz four-core Power Mac G5 in Speedmark. In some tests, such as iTunes and iMovie HD, the differences were minimal, but in our HandBrake test it took the four-core Power Mac G5 nearly 18 times longer to encode the movie than the recommended Mac Pro configuration. For those wanting more comparisons, here’s a link to
many more Speedmark 5 results.
We’re hoping to get in our 3.2GHz Mac Pro sometime next week, and also hope to have different graphics cards to test in our new Mac Pros. We’ll get those results up as soon as we can. Also look for our full review of the new Mac Pro soon.
This article was updated on January 25 to correct data-transcription errors for the Power Mac G5’s results in our Finder and HandBrake tests.