Mac mini Core Duo 512MB vs 1GB--How Fast Is It?

We recently applied our speed tests to our 1.66GHz Mac mini Core Duo with 512MB of RAM in a shootout among our upgraded Power Mac G4 Cube (that the Mac mini Core Duo replaced) and our 20" 2GHz iMac Core Duo. One question left hanging was, should one upgrade the base Mac mini Core Duo's RAM from 512MB to 1GB or even to 2GB?

We tested the RAM requirements of our original 1.25GHz Mac mini G4 and came to the conclusion that 256MB was easily sufficient for an entry-level or general computer user. We also saw no difference in application speed during weeks of use with 256MB or 512MB of RAM. A stress test consisting of launching and running six popular applications simultaneously showed no improvement when the 256MB RAM module was replaced by a 512MB module. Of course, we did not run any professional video or digital image applications which run better with a vast amount of RAM because the Mac mini G4 was not targeted at those users. But what about the Mac mini Core Duo that needs Rosetta to run many application?

The Mac mini Core Duo is still targeted at the entry-level and general computer user, especially those who already have a display, keyboard and mouse that they want to continue using. The Mac mini Core Duo and Solo have gone upscale in price relative to the Mac mini G4 by $100 but with much more capability and speed. There is better value, in our opinion, in the iMac Core Duos with faster Intel CPUs and much better graphics chips, so we expect that the Mac mini users will continue to be entry level or general computer users. Our Mac mini Core Duo replaced our upgraded G4 Cube for daily general computing.

First, we ran our normal suite of speed tests (see our Speed FAQ) on our Mac mini Core Duo with 512MB (two matched 256MB modules) of RAM and again with 1GB (two matched 512MB modules). Here are the results:

Table 1. Speed Test--1.66GHz Mac mini Core Duo with 512MB vs 1GB RAM (average* scores or secs).
Test
Mac miniCore Duo
512MB RAM
Mac mini Core Duo
1GB RAM
Benchmarks:
AltiVec Fractal
3.3
3.3
Let1kWindowsBloom (sec)
19
19
Cinebench 9.5 - Rendering
256/464+
258/474+
- OpenGL Hardware Lighting
546
549
Xbench 1.2 - CPU
64
64
- Threading
166
166
- Memory
97
100
- Quartz Graphics
60
60
- OpenGL Graphics
183
183
- User Interface Graphics
n/v
n/v
- Hard Disk
30
30
"Real World" Tests:
Startup (sec)
25
25
Shutdown (sec)
8
9
Duplicate 500MB folder of files (sec)
48
48
Duplicate 700MB folder of 3600 files (sec)
68
66
Scroll File in Adobe Reader 5.1 (sec)
35
35
Quake III Arena (frames/sec)
44
44
*Average of 3 runs. +Using both cores. n/v=not valid

The scores or times are virtually identical, indicating that the extra 512MB of RAM did not help with the performance of these tests.

Next, we ran our stress test. We ran it with 256MB (one module removed), 512MB and 1GB of RAM. Six applications were simultaneously opened: Safari, Mail, iTunes, QuickTime Player, Microsoft Word and Microsoft Excel. We played a Photoshop TV episode in iTunes and played Steve Jobs' Macworld 2006 keynote streamed from an Apple server in QuickTime Player. Here are the RAM allocations shown by Activity Monitor:

Table 2. Mac mini RAM Allocation under Stress
RAM Allocation
256MB
512MB
1GB
.
Wired
125
146
178
Active
72
226
225
Inactive
36
114
512
Free
3
6
91

Subjectively, each of the applications seemed to work fine with 512MB of RAM and all of the applications open and seemed to work no better with 1GB. However, at 256MB the slow-down was very clear. Basically, the Mac mini Core Duo was unusable at 256MB--applications took forever to launch and switch between and startup took 62 seconds as opposed just 25 seconds at 512MB and 1GB.

Then we ran Xbench 1.2 to see if it showed any better performance with 1GB than 512MB or 256MB of RAM. Again, the Mac mini Core Duo performed extremely sluggishly with 256MB, taking 15 minutes and 2 seconds to complete one run. Xbench 1.2 took and average of 118 seconds with 512MB and 114 seconds with 1GB, a slight but not significant difference. Here are the Xbench scores:

Table 3. Xbench 1.2 Scores* with Apps Open
Xbench Test
256MB
512MB
1GB
.
CPU
63
63
62
Threading
136
136
137
Memory
2
92
94
Quartz Graphics
46
54
57
OpenGL Graphics
148
174
165
User Interface
n/v
n/v
n/v
Hard Disk
24
28
28
*Average of 3 runs. n/v=not valid

Generally, all of the scores are lower with the six applications open (Table 3 vs Table 1) showing the effect of the open applications on the performance measured by Xbench. The memory scores show the most difference, with 256MB getting only a 2 versus 92 and 94 for 512MB and 1GB respectively. There is very little, if any, real difference between the Xbench scores using 512MB or 1GB of RAM.

Clearly, the Mac mini Core Duo can not run on 256MB of RAM, unlike the Mac mini G4 which did well with 256MB on our tests and during general use. This may be due to the Core Duo architecture requiring Rosetta to run some of the applications. It is also clear to us that the benefit, if any, of 1GB of RAM over the standard 512MB is small. Most people can save $100 and stick with the Mac mini Core Duo's base 512MB of RAM. Those who are likely to occasionally work with very large images on a Mac mini, like us, may benefit from the upgrade and may wish to consider it. [Bill Fox]



Saw something? Send a tip

The archive ran on reader tips. What did you see, where, and do you want the credit?

Read by the editor. Never published without your say.

More in Reviews · This month in the archive