MacInTouch Special Reports

 Ads     E-mail     Find     Home     Resources     Sitemap 


 

Macintosh Performance Comparison:
eMac G4, iBook G4, iMac G5, Mac Mini, PowerBook G4

  1. Introduction
  2. Xbench
  3. Cinebench
  4. File Duplication
  5. MPEG4 Conversion
  6. Conclusions

  7. Pros and Cons
  8. Links
  9. Updates


Introduction

With several of Apple's low-midrange computers in hand, we decided to compare performance using our standard benchmark test suite, finding surprising speed in the new eMac and iBook G4 models. For perspective, we compared results with iMac G5 and iBook G3 tests, as well as a RAID-equipped Power Mac G5. (You can compare these results to those of additional systems, such as the Power Mac G4/450 and G4/800, in our iMac G5 Benchmarks page.)

When the exciting, low-cost Mac Mini arrived in January, we ran benchmarks on a 1.25GHz model we bought and updated our report. Rob White subsequently supplied Xbench results for a 1.42GHz model, and we later got an explanation for G5 disk performance problems.

When Apple introduced the new PowerBook G4 laptops on January 31, 2005 (lowering prices), we ordered the 15" model, which won widespread praise in the past. More expensive than the other systems here, the PowerBook offers a very interesting contrast to the iMac G5, combining a very fast G4 processor with FireWire 800 support and a new graphics controller that can drive an Apple flat-panel monitor just like the iMac G5's.

Test Systems

We used Mac OS X 10.3 "Panther" for these tests. In general, we tested at millions of colors and default settings for most things. The processor setting - located in Preferences > Energy Saver > Options - can make a big difference in computers that offer the option: laptop computers and G5 systems. We suggest keeping it at the default "Automatic" setting in general, but you can get more performance by setting it to "Highest", at the cost of more heat (and fan noise) and shorter battery life.

We evaluated journaling vs. non-journaled disk volumes and found surprisingly little difference between the two, although it probably should be disabled for the highest performance. We also found little difference in having AirPort turned on or off, and we typically keep a firewall enabled, which should have little effect on performance.

Running other applications, however, can affect test results, so we tried to run only the benchmark application by itself.

Xbench is somewhat inconsistent from run to run, so we averaged at least three results. We did the same with OpenGL tests in Cinebench.


Xbench Results

Xbench Test Results


Xbench, which attempts to test all the major Mac hardware systems, shows several things clearly in a graph of test results:

  1. The Energy Saver processor option makes a huge difference in performance
  2. The inexpensive iBook G4, eMac G4 and Mac Mini are surprisingly fast - at least as fast as the iMac G5 in its normal mode.
The eMac G4 easily exceeds both iBooks and pushes the iMac G5 very hard, despite the iMac's fancier G5 processor, thanks to the eMac's superior disk performance.

The eMac vs. Mini comparison is especially interesting. The Mini is hampered by a slow disk drive (see below), yet the 1.25GHz Mini matches the eMac overall, thanks to better scores for Quartz graphics and "user interface" performance. The 1.42GHz Mini is a little faster than the 1.25GHz model for its small price premium.

The PowerBook G4/1.67GHz has an advantage over the slower Mini and eMac when set to "Highest" processor mode, but lags behind them in "Automatic" processor mode.

Comparing iMac G5 vs. PowerBook G4 is interesting. The PowerBook actually has higher CPU and Thread Xbench scores! However, the iMac G5 wins in memory speed (a major G5 design feature), graphics and internal disk performance, especially in "Highest" processor mode.

 

Cinebench Results

Cinebench Test Results

Cinebench rankings of graphics performance put our Macs into two distinct categories: expensive and fast vs. cheaper and slower.

The iMac G5 and Power Mac G5 led the pack, with the PowerBook G4 hot on their heels. The other Macs - iBook, Mini and eMac - all share Radeon 9200 graphics controllers, and their test results were similar, well off the top pace. (Disk speed obviously wasn't a factor.)



File Duplication

File Duplication Test Results


One of our standard "real-world" tests of disk performance is to duplicate a large folder in the Finder. The folder contains ten copies of a 20.7-MByte file, for a folder size of 207.5 MBytes. We duplicate it several times and use a stopwatch to measure how long it takes.

We found these test results hard to believe at first: The eMac G4 outperformed the iMac G5. We double-checked results, but all the numbers appeared valid. Then we checked the hard drive models: This 160GB eMac uses a parallel version of the same Seagate drive used in the iMac G5. Aside from the interface, they should perform the same.

We then realized that these results echoed our earlier comparison of a Power Mac G4/450 with the iMac G5. The much older G4, updated with an equivalent drive, had disk performance as fast as the brand-new iMac G5 (and that's without using RAID to boost the G4 further).

What's going on? We finally got an explanation from a former Apple engineer (who also noted understaffing issues in Apple engineering). To summarize, Apple apparently made some design errors in low-level hardware priorities for the custom controller chips, starving the I/O system for memory bandwidth to feed the fast G5 processors. As a result, the much-slower G4 systems actually perform better in disk operations.

Another controller problem is said to be responsible for poor USB 2.0 disk performance with Macs, which, like FireWire, apparently runs faster on some Windows hardware.

In any case, disk performance is a clear win for the eMac, one of the best things about this computer. The only way to beat the eMac's disk performance is with a Power Mac G5 set up with a striped dual-disk RAID (or, something we haven't tested, a fast storage system connected to a PowerBook via FireWire 800.)

As expected, with its 2.5-inch laptop drive, the Mac Mini lagged behind other desktop Macs with slower disk performance, equivalent to the iBook G4's. Unfortunately, there's no good way around this limitation without FireWire 800 in the Mini. FireWire 400 is only a bit faster with a high-performance hard drive, and a fast hard disk drops to a terribly slow pace on USB 2.0 (worse than the internal notebook drive).

The Mini is absolutely not designed for internal access by customers, but (with two decades of Mac hardware experience) we opened it up and just managed to swap out the Apple hard drive for the fastest 2.5-inch drive we could buy, a Hitachi Travelstar 7K60, which cost $180. (After our experience, we strongly recommend against doing hard drive upgrades in the Mini, unless you're a very accomplished and confident hardware hacker. See more details in our Mac Mini Review.)

The Travelstar 7K60 upgrade did improve Mini disk performance, as shown in the chart, to iMac G5 levels. This still lags the eMac, as well as Power Mac RAID systems or fast FireWire 800 disks attached to larger PowerBooks. So, it doesn't make sense to buy a Mini for a disk-intensive application, but it's fine for all sorts of office tasks and other applications, such as domain name servers, that aren't disk-intensive.

The iMac G5 also suffers from the inability to host multiple drives on a fast interface, but at least the iMac accomodates a single, fast 3.5-inch internal drive. The iMac runs the identical drive a little slower than the eMac, thanks to the controller issues, but it should be fine for all normal uses.

In this real-world test, the PowerBook G4's internal, 80GB Fujitsu drive showed performance similar to the 80GB Toshiba drives we tested in the Mini and iBook, which is nothing special - the 60GB Hitachi Travelstar 7K60 we put into the Mini was substantially quicker.

Plug a decent FireWire 800 disk into the PowerBook G4, however, and you've got a whole new ballgame. The PowerBook is suddenly faster than everything but the eMac and Power Mac RAID system, easily beating the iMac G5 and Mini. If you connect two FireWire 800 drives and combine them into a mirrored RAID with Disk Utility, you've got disk performance that outruns every other Mac, except a Power Mac with internal, striped RAID. This is the kind of performance that's really going to accelerate real-world work, and it's a compelling reason to buy a 15"/17" PowerBook over a Mini, iMac G5, iBook or 12" PowerBook. (We found minimal slowdowns in disk performance switching Energy Saver from Highest to Automatic processor mode.)

 

MPEG4 Conversion

MPEG4 Encoding Test Results


Using QuickTime Pro and QuickTime Player, we export a high-quality 50-second DV file to MPEG-4 format. Source and target files are on the hard drive. We use the standard "Default" settings.

This is a good real-world test of system performance, and the results are surprising: Right out of the box, the eMac, Mini and lowly iBook G4 handily outperform the iMac G5 at Apple's standard settings!

If you change Apple's Energy Saver options to get "Highest" processor performance, the iMac G5 finally beats its cheaper siblings, but there must be some reason that's not the default, and clock speed alone should give the iMac a big advantage. In return for pushing the iMac, you're likely to get more fan noise at the very least.

Apple is using the G5's special "slewing" feature to reduce heat and power drain, and the result is a real bottleneck. In real, everyday applications, Apple's cheapest G4 systems may pace the fancy G5 models at a half or third of its cost.

Apple's G4 portables face the same issue as the G5 systems, for the sake of longer battery life, running slowly in their normal "Automatic" mode. Crank the processor mode up to "Highest", however, and you'll pump out movies faster, even on a lowly iBook, than you could using an iMac G5 in its standard mode of operation!


Conclusions

The iMac G5 is a wonderful system, and we'd rather pay a few hundred dollars over the cost of an eMac to get one, but all the Apple hype about the G5 falls a little short when you see the low-cost eMac, with its slower G4 processor, pushing the iMac G5 in performance. The eMac is actually faster in several real-world situations, and it looks like low-level hardware design is responsible for eMac advantages in disk performance.

Adding the Mac Mini to the mix really changes the buying equation. For far less money than an iMac or even an eMac costs, you get excellent performance, silent operation and the ability to drive a big beautiful monitor of your own choosing, a critical feature missing from all but the Power Mac and big PowerBooks. The Mini's one weakness is disk performance, which may make the eMac a better choice for video, database and multimedia work, but it shouldn't be an issue in too many applications. For general home or office use, the Mini is perfect.

If you want the ultimate in performance - or multiple large screens - the Power Mac G5 is the way to go, although we have some concerns about reliability with the liquid-cooled 2.5GHz model and would probably stick with 1.8- or 2.0-GHz systems. The ability to hold dual internal hard drives for a RAID configuration easily overcomes the controller performance problems.

PowerBooks are nice, but pricy. The one advantage you get for the extra price of the 12" PowerBook is the ability to drive a larger external screen in dual-display mode (up to 2048x1536), although the built-in screen has the same 768x1024 resolution as the iBook.

The 15" PowerBook is an ideal mobile machine, and it can drive a big external screen and use FireWire 800 to get outstanding disk performance in a compact, quiet package. This lovely laptop costs about four times what a Mac Mini costs, however, making it an expensive option for its portability and performance.

The 17" PowerBook strikes us as an expensive alternative to the iMac with better portability and battery power. It's basically an oversized 15" PowerBook.

One final factor is the G5's support for 64-bit processing, which is supposed to get a boost with this year's Mac OS X 10.4 "Tiger" release. Theoretically, this may be an advantage for the iMac and Power Mac G5 models, but the real-world advantage for general applications is questionable at this point. For specialized scientific or business processing, it could give G5 Macs an advantage over other platforms.


By Ric Ford



Pros and Cons

Pros Cons
Mac Mini fast CPU/memory, very quiet and compact, inexpensive disk performance lags all but iBook; no FireWire 800 or audio input
eMac G4 surprisingly fast (especially disk); audio line input; screen resolution goes fairly high bigger, heavier and louder than others; screen not as nice
iBook G4 fast for a laptop, very compact and quiet, appealing price limited screen resolution (768x1024) without unsupported hack, no audio input, not as fast as desktop systems
iMac G5 compact, very fast CPU/memory, exceptional access to internals, beautiful screen more expensive than eMac/iBook/Mini, disk performance slower than expected, lacks FireWire 800
PowerBook G4 great screen and external display support; excellent performance; FireWire 800 offers dramatic disk speed boost, audio input and PC Card/CardBus; quiet expensive; internal disk not very fast





Document History

 


 Ads     E-mail     Find     Home     Resources     Sitemap

Copyright 2004-2005 by MacInTouch, Inc. All rights reserved.



click for more info