Lab Report: New build-to-order iMacs impress with speed

When Apple released new iMacs in early May, the company also made available new build-to-order (BTO) options in addition to the standard-configuration models. Macworld Lab tested two iMacs with BTO processor upgrades that offer faster speeds and more processing power, and the results show that the additional cost for the upgrades are worth it—if you run software that takes advantage of the technology.

Processor upgrades are optional for two of the four standard-configuration iMac models. For the $1499 21.5-inch 2.7GHz Core i5 quad-core model, an upgrade to a 2.8GHz Core i7 quad-core processor is available. For the $1999 27-inch 3.1GHz Core i5 quad-core iMac, you can upgrade to a 3.4GHz Core i7 quad-core processor. Each upgrade adds $200 to the price of the respective standard-configuration computer.

The standard-configuration iMacs use processors that do not use Intel’s Hyper-Threading technology, but the Core i7 upgrades do support the technology. Hyper-Threading allows two processing threads to run on a single core. In the case of a quad-core processor, Hyper-Threading would present the operating system with eight virtual cores. This can allow for faster processing. The only way to get Hyper-Threading on the new iMacs is to buy the BTO Core i7 processor option.

However, many applications do not make efficient use of multiple cores, so going from four cores to eight virtual ones doesn’t make a lick of difference. Some applications, however, can and do make use of all of those cores. Processor-intensive applications such as HandBrake, Cinema 4D, and Mathematica benefit greatly from Hyper-Threading.

BTO iMacs (Mid-2011) benchmarks

Longer bars are better. Blue bars in italics represent reference systems. Macworld Lab testing by James Galbraith, William Wang, and Mauricio Grijalva.

Whether the upgrades are worth the additional cost depends on what you plan to do with your iMac. Our system performance test suite, Speedmark 6.5, produced results that show that the 21.5-inch iMac with a BTO 2.8GHz Core i7 processor is about 7 percent faster than the standard 21.5-inch 2.7GHz Core i5 iMac. The 27-inch iMac with a BTO 3.4GHz Core i7 processor is about 11 percent faster than the standard 27-inch 3.1GHz Core i5 iMac.

In individual tasks, the 21.5-inch 2.8GHz Core i7 iMac was more than 15 percent faster than the standard 21.5-inch 2.7GHz Core i5 iMac when importing a camera archive into iMovie, 18 percent faster in our HandBrake encoding test, 22 percent faster in MathematicaMark, and 25 percent faster in the Cinebench CPU test. The BTO 21.5-inch iMac was 11 percent faster overall than the $1199 21.5-inch 2.5GHz Core i5 iMac, and 21 percent faster in our iMovie import test, 24 percent faster in HandBrake, 30 percent faster in Cinebench’s CPU test, and 31 percent faster in MathematicaMark.

The BTO 21.5-inch iMac was also 5 percent faster overall than the standard $1999 27-inch 3.1GHz Core i5 iMac, and 7 percent faster in our HandBrake test, 9 percent faster in our iMovie import test, 12 percent faster in MathematicaMark, and 15 percent faster in the Cinebench CPU test.

The 27-inch iMac with a 3.4GHz Core i7 processor was 5 percent faster overall than the BTO 21.5-inch iMac, and it was 8 percent faster in our iMovie import test, 15 percent faster at HandBrake, and 17 percent faster at both Cinebench CPU and MathematicaMark.

Speedmark 6.5 results showed the BTO 27-inch iMac to be 11 percent faster than the standard $1999 3.1GHz Core i5 iMac. The 3.4GHz Core i7 iMac was 17 percent faster in iMovie, 21 percent faster in HandBrake, 29 percent faster in the Cinebench CPU test, and scored 31 percent higher than the 3.1GHz i5 iMac in MathematicaMark.

The new standard 27-inch 3.1GHz Core i5 iMac performed very similarly to a BTO model from last year, a 27-inch 2.93GHz Core i7 quad-core iMac.

Our Speedmark 6.5 test results page shows how the BTO iMacs compare to other current and discontinued Macs.

Speedmark 6.5 individual application test results:
New BTO iMacs (Mid 2011)

Duplicate
1GB
File
Zip
2GB
folder
Unzip
2GB
folder
Pages '09
Open
Word Doc
21.5-inch iMac 2.8GHz Core i7 (BTO) 20 130 45 65
27-inch iMac 3.4GHz Core i7 (BTO) 20 130 45 63
21.5-inch iMac 2.7GHz Core i5 (Mid 2011) 20 134 47 65
27-inch iMac 3.1GHz Core i5 (Mid 2011) 20 144 48 70
27-inch iMac 2.93GHz Core i7 (BTO, Mid 2010) 17 157 36 71
27-inch iMac 3.6GHz Core i5 (BTO, Mid 2010) 17 140 36 65
Mac Pro 3.33GHz Xeon Westmere six-core (BTO, Mid 2010) 18 148 26 62

Results are in seconds. Lower results are better. Reference models in italics. Best result in bold.

Speedmark 6.5 individual application test results:
New BTO iMacs (Mid 2011)

iTunes 10
AAC to
MP3
encode
Import
movie
archive to
iMovie '09
iMovie '09
export to
iTunes 10
for iPhone
Call of
Duty 4
framerate
21.5-inch iMac 2.8GHz Core i7 (BTO) 72 60 49 88.1
27-inch iMac 3.4GHz Core i7 (BTO) 68 55 46 88.6
21.5-inch iMac 2.7GHz Core i5 (Mid 2011)
74 71 49 87.6
27-inch iMac 3.1GHz Core i5 (Mid 2011)
74 66 51 88.5
27-inch iMac 2.93GHz Core i7 (BTO, Mid 2010) 92 61 55 88.3
27-inch iMac 3.6GHz Core i5 (BTO, Mid 2010) 80 90 61 79.0
Mac Pro 3.33GHz Xeon Westmere six-core
(BTO, Mid 2010)
81 48 48 88.3

Call of Duty 4 results are based on framerate; higher results are better. All other test results in the above chart are in seconds; lower results are better. References models in italics. Best result in bold.

Speedmark 6.5 individual application test results:
New BTO iMacs (Mid 2011)

iPhoto '09
200 JPEG
import
Photoshop
CS5 action
HandBrake
0.9.4
encode
Cinebench
R11.5
graphics
21.5-inch 2.8GHz iMac Core i7  (BTO) 35 50 194 42.6
27-inch 3.4GHz iMac Core i7 (BTO) 33 51 165 43.2
21.5-inch 2.7GHz iMac Core i5 (Mid 2011)
33 52 236 41.6
27-inch iMac 3.1GHz Core i5 (Mid 2011)
28 53 209 38.3
27-inch iMac 2.93GHz Core i7 (BTO, Mid 2010) 29 53 208 34.8
27-inch iMac 3.6GHz Core i5 (BTO, Mid 2010) 33 55 340 31.0
Mac Pro 3.33GHz Xeon Westmere six-core
(BTO, Mid 2010)
27 55 144 34

CineBench R11.5 Graphics results are a score; higher results are better. All other test results in the above chart are in seconds; lower results are better. Reference models in italics. Best result in bold.

Speedmark 6.5 individual application test results:
New BTO iMacs (Mid 2011)

Cine-
bench
R11.5
CPU
Mathe-
matica
Mark 7
Parallels
World-
Bench
Multitask
test
Aperture
3 import
and
process
Multi-
tasking
21.5-inch iMac 2.8GHz Core i7 (BTO) 72 13.1 254 101 59
27-inch iMac 3.4GHz Core i7 (BTO) 60 15.3 226 100 55
21.5-inch 2.7GHz iMac Core i5 (Mid 2011)
96 10.7 262 104 61
27-inch iMac 3.1GHz Core i5 (Mid 2011)
85 11.7 247 107 58
27-inch iMac 2.93GHz Core i7 (BTO, Mid 2010) 77 12.5 282 108 62
27-inch iMac 3.6GHz Core i5 (BTO, Mid 2010) 130 7.0 293 109 72
Mac Pro 3.33GHz Xeon Westmere six-core
(BTO, Mid 2010)
48 19.2 253 101 59

MathematicaMark 7 results are scores; higher results are better. All other test results in the above chart are in seconds; lower results are better. Reference models in italics. Best result in bold.

How we tested. Speedmark 6.5 scores are relative to those of a 2.4GHz Core 2 Duo Mac mini (Mid 2010) with 2GB of RAM, which is assigned a score of 100. We duplicated a 1GB file, created a Zip archive in the Finder from the two 1GB files and then unzipped it. We converted 135 minutes of AAC audio files to MP3 using iTunes’ High Quality setting. In iMovie ’09, we imported a camera archive and exported it to iTunes using the Mobile Devices setting. We ran a Timedemo at 1024-by-768 with 4X anti-aliasing on in Call of Duty 4. We imported 200 JPEGs into iPhoto '09. The Photoshop Suite test is a set of 23 scripted tasks using a 50MB file. Photoshop’s memory was set to 70 percent and History was set to Minimum. For our multitasking test, we timed the Photoshop test again, but with the iTunes MP3 encoding and file compression tests running in the background. We used Handbrake to encode four chapters from a DVD previously ripped to the hard drive to H.264. We recorded how long it took to render a scene with multiprocessors in Cinebench and ran that application's OpenGL, frames per second test. We ran the Evaluate Notebook test in MathematicaMark 7. We ran the WorldBench 6 multitasking test on a Parallels 6 VM running Windows 7 Professional. We timed the import and processing time for 200 photos in Aperture.—Macworld Lab testing by James Galbraith, William Wang, and Mauricio Grijalva

Macworld Lab is working on more performance results with the BTO iMacs, including tests with solid-state drives. Have ideas for more testing? Let us know in the comments section below.

[James Galbraith is Macworld’s lab director.]

Comments