Apple's Sandy Bridge-Based Mac Mini Review
Cross-Platform Tests / Benchmarking
Our Test Methodologies: As the Mac mini uses Mac OS X Lion (10.7), we weren't able to use our normal arsenal of Windows-based benchmarks. So we devised a number of new tests that we could use to compare the machine against a number of other Macs and Windows systems. These tests are broken up into two sections: The first set (below and on the next page) are cross-platform tests, where the same workload was run on both the Mac and Windows systems. The second set consists of Mac-only tests. All tests were run several times on each system to ensure consistency. The comparison systems are as follows:
-
A two-year old, 24-inch iMac. The iMac's config is: 2.8GHz Intel Core 2 Extreme X7900, 4GB 800MHz DDR2 SDRAM, ATI Radeon HD 2600 Pro, 750GB 7,200-rpm hard drive, and running Mac OS X 10.6.4. When this iMac was released, it represented close to the top-end of available configs, and it is still considered reasonably powerful--especially when compared against a laptop.
-
One of the two Windows comparison laptops has this config: 2.13GHz Intel Core i3-330M, 4GB 1,066MHz DDR3 SDRAM, Integrated GMA HD, 320GB 5,400-rpm hard drive, and running Windows 7 Home Premium (64-bit).
-
The other Windows comparison laptop has this config: 1.2GHz Intel Core i3-330UM, 3GB 1,066MHz DDR3 SDRAM, Intel GMA HD and ATI Radeon HD 5450 (512MB) switchable graphics, 500GB 7,200-rpm hard drive, and running Windows 7 Home Premium (64-bit).
 |
|
To touch on gaming
As always, comparing an Apple machine to a Windows-based machine isn't a true Apples to Apples comparison (no pun intended). We compared the Half-Life 2: Episode 2 test to a slew of other Windows based machines. There are two main things to take away from this, comparisons aside. First, althouhg it is not representd in the graph, it's was nice to find that this machine was able to run this game at 1920x1200 with "High" details and still hit 30-31 frames per second. We maxed out the screen resolution on our test panel (which is also the highest resolution supported through the HDMI socket), and HL2:EP2 ran well.
Secondly, this is no gaming machine. Not even close, despite handling this title with ease. If you tried to run new titles, you'd probably have to chuckle at the results. But Apple never engineered this to be a cutting-edge gaming machine, and given just how few titles actually run on OS X at all, you probably shouldn't even be considering this machine if gaming is a top priority. Windows-based rigs still provide the best gaming experience. The upside is that some light/medium-duty gaming isn't impossible here, but anything newer than a year or so is probably too intense for that integrated graphics processor to handle.
Maxon's Cinebench R11.5 benchmark is based on Maxon's Cinema 4D software used for 3D content creation chores and it tests both the CPU and GPU in separate benchmark runs. On the CPU side, Cinebench renders a photorealistic 3D scene by tapping into up to 64 processing threads to process more than 300,000 total polygons; while the GPU benchmark measures graphics performance by manipulating nearly 1 million polygons and huge amounts of textures.
Secondly, this is no gaming machine. Not even close, despite handling this title with ease. If you tried to run new titles, you'd probably have to chuckle at the results. But Apple never engineered this to be a cutting-edge gaming machine, and given just how few titles actually run on OS X at all, you probably shouldn't even be considering this machine if gaming is a top priority. Windows-based rigs still provide the best gaming experience. The upside is that some light/medium-duty gaming isn't impossible here, but anything newer than a year or so is probably too intense for that integrated graphics processor to handle.
 |
|
Maxon's Cinebench R11.5 benchmark is based on Maxon's Cinema 4D software used for 3D content creation chores and it tests both the CPU and GPU in separate benchmark runs. On the CPU side, Cinebench renders a photorealistic 3D scene by tapping into up to 64 processing threads to process more than 300,000 total polygons; while the GPU benchmark measures graphics performance by manipulating nearly 1 million polygons and huge amounts of textures.
Not that the OpenGL score for the mini is actually (slightly) lower than the 13" MacBook Air we reviewed in November of 2010, due to the Air's NVIDIA GPU. But having a lowly Intel IGP and just 2GB of shared system RAM doesn't help matters. The CPU score was obviously a bit higher, but that's what having a Core i5 will do for you. The reality is this: the Mac mini is no powerhouse when it comes to editing serious photos/videos or rendering. If you're a creative pro, you need more horsepower (and more flexibility) than is offered here. That's why the Mac Pro exists. But in a pinch, this is perfectly suitable for the quick Photoshop touch-up and iMovie edit. We experienced no notable lag when doing basic media editing, and the average consumer probably won't stress the limits in these areas, anyway.
