Benchmarking the Mac mini 2.26 (Part 2)

icerabbit on April 12th, 2009

[ Part 1 is posted here ]

Here at icerabbit HQ we have been a little too busy recently to focus on testing the new Mac mini 2009 2.26GHz with other hard drives, after we upgraded the memory. Last night I finally gave the mini a nice place on my desk under the 20″ Cinema Display to which it is hooked up.

My original plan has been to use the Mac mini with a faster 3.5″ external hard drive, rather than upgrading the internal drive, which is rather tricky. If you like detailed instructions to perform such an upgrade, look no further than this iFixit guide.

Noting that I haven’t switched to the mini as my primary machine yet; so far, I have been very satisfied with the speed of its 2.5″ Fujitsu …. hard drive. The combination of this drive + the mini is more responsive than many macs with 5400 rpm drives I have seen over the years. So, I haven’t special ordered an anno 2009 zippier 3.5″ 7200rpm drive yet for the ministack. Right now it looks like I will use the ministack for secondary mass storage with a regular drive.

I still want to do some tests using the newertech miniStack v3 with some older drives as well as a couple retail external drives from Maxtor (3.5″USB2 & FW400) and Seagate (2.5″USB2). While I agree that the internal SATA bus is the fastest connection, it doesn’t mean the internal hard drive is the fastest by default. Regardless of connection method, the internal hard drive is still the slowest component in the computer (apart from the optical drive) and so an external hard drive could be faster.

Even if I don’t find something faster than the internal drive in the house it will be a fun exercise to see how long it takes to back up the mac mini using USB2, FireWire400 and FireWire800. What external bench scores will be? Which will be the faster drive? Which applicaion will be faster? SuperDuper or CarbonCopyCloner? And what about temperatures? Does all this hard drive activity kick the fans in overdrive, or not? Click through for the findings.

Preparations

I let OSX update do its thing and then did a disk repair. To my surprise that took a while. Four and half minutes to read the permissions database, followed by 1 minute to do the actual repair.  Hard drive info: 16.8GB used, 146000 folders and 546500 files (rounded)

All hard drives were erased / formatted using Disk Utility to OSX Extended Journaled each time prior to using SuperDuper & Carbon Copy Cloner.

Cloning

Maxtor OneTouch4 Plus

External 500GB Maxtor OneTouch 4 Plus USB2 & FW400

SuperDuper Clone Backup

Superduper – All Files via USB2.Preparation completed in 20 secs. Then starts copying files. Effective copy speed varies greatly. It first hovers between 2.75MB/s and 3.75MB/s, then jumps to 9+MB/s, slows to 8+, 6+, 5+. You get the idea. 5GB is transfered in 16 minutes. 10GB in 27 minutes. In all it takes 49 minutes to back up 16.8GB completely.

SuperDuper Backup – All Files via FireWire400 has the preparation done in 13 secs and then starts copying files. Effective copy speed varies greatly as well depending on the files being copied. 5GB done in 16 minutes. 10GB in 26 minutes. In all it takes 46min 28s to back up 16.8GB.

Conclusion: using SuperDuper, the 16+ GB backup did not show any significant difference between using USB & FW towards the Maxtor OneTouch4 drive.

Carbon Copy Cloner Backup

Disk-to-Disk clone via FireWire takes off pretty good. 1GB in 1:30. 2GB in 3:53. 5GB in 9:28. 10GB in 19:15 and the entire backup completed in 29:15. Very impressive compared to SuperDuper.

Disk-to-Disk clone via USB is slower off the mark compared to FireWire. 1GB in 1:45. 2GB in 4:20. 5GB in 10:01. 10GB in 20:50 and the entire backup was  completed in 31:03. Not quite as fast as FireWire, but still way better than using SuperDuper through both.

Seagate FreeAgent Go 250

Seagate FreeAgent Go 250 USB 2

Starting with Carbon Copy Cloner. Missed the 1GB time, but it was around 2min. 2GB in 4:18. 5GB in 10:32. 10GB in 21:28. Task completed in 31:56. A bit slower, but as a whole about on par with the Maxtor drive.

SuperDuper? I think we can safely assume it is going to be >45-50 minutes and I’m taking the liberty to skip that test.

 

Newer Technology ministack v3 FW800

I will actually need to update this post in the upcoming days as I don’t have an empty spare SATA drive handy right this minute. 

Benchmarks

 

Test results for my mac Mini 2009 2.26 Ghz 4GB RAM using the

FUJITSU MHZ2120BH G1 internal 2.5″ HDD

GeekBench: 3018

 Geekbench Mac mini 2.26 GHz

Xbench scores 130.57 & 128.7 with hard drive scores of 48.9 & 48.5 respectively

Xbench Mac mini 2.26 results

 

Maxtor OneTouch4 Plus

Booted via USB2. From option selection to full desktop: 21 & 23 seconds
Xbench scores: 115.08 & 114.6 with hard drive scores of 36.22 & 37.13

Booted via FireWire400: 28 & 25 seconds (which surprised me to be slower)
Xbench scores: 128.9 & 124.7 with hard drive scores of 45.01 & 44.98

Boot times seem to edge out for USB2, but clearly connecting to the Maxtor via USB adversely affects the overall and hard drive score in Geekbench; where via FireWire it keeps up with the internal hard drive.

Seagate FreeAgent Go 250

Booting was strangely slow at 1:23. The second time a more normal 35s from Option to completed desktop.
Xbench scores: 107.54 & 102:16 with hard drive scores of  30.49 & 29:21

Clearly this drive performs less well than the Maxtor.

Newer Technology ministack v3 FW800

Coming soon.

Temperature & Noise

Temperature was not an issue. During testing the CPU temperature would rise from 50C to 63C and I noted a max of 71C at one point, but the fan(s) remained within plus or minus 5 rpm of 1500rpm and quiet.

The internal hard drive was near silent. Its temperature rose from 46C to 52C. Now sits idle at 42C with an ambient temperature of 65F.

3 Responses to “Benchmarking the Mac mini 2.26 (Part 2)”

  1. Great testing.
    Looking forward to your test with the ministack v3.

  2. Nice writeup and illustrations. I’m comparing FW and USB external drives on my Mini right now, so the test case and numbers prove quite helpful.

  3. Thank you, Matthew.
    Glad you found the numbers useful. I didn’t unpack the ministack completely last year and it has been sitting idle on the shelf. I must make some time to put a drive in it and see how it works.

Trackbacks/Pingbacks

Leave a comment - Voeg een reactie toe