Just like early reports on the new MacBook Air released yesterday, the new Mac mini is also showing significant improvements in raw performance over its predecessor machines. Mac mini web hosting company Macminicolo.net has obtained both standard and server models of the new line and put them through some preliminary benchmarking using Geekbench.
/article-new/2011/07/mac_mini_2011_geekbench1.jpg?lossy "mac_mini_2011_geekbench")
The testing reveals significant performance improvements for the Mac mini, with the entry-level 2.3 GHz Core i5 model showing an over 50% jump in Geekbench scoring over earlier generations and the 2.0 GHz Core i7 server model showing even more impressive results with its Geekbench score coming in at well over double that of its predecessor.
As we noted on yesterday's MacBook Air report, Geekbench focuses on raw processor and memory performance, and real-world performance will also depend on other aspects of a system including graphics capabilities and data storage components.
/article-new/2011/07/mac_mini_2011_unbox.jpg?lossy)
The report also offers a few unboxing photos and some shots of the initial setup process in which the Mac mini server model offers to migrate from an existing server setup. Also noted is the fact that the server edition includes iLife apps, as the server portion of Lion is simply an add-on for the basic OS X Lion. Under Mac OS X Snow Leopard, iLife apps were not included on server machines.
Top Rated Comments
Works no problem for me....
And very strangely, upgrading from one 500 GB drive to one 750 GB drive in the desktop model costs $150. Upgrading the server model from dual 500 GB's to dual 750 GBs costs only $100... Uh, what? :confused:
Nope, for multithreaded CPU-intensive apps (like video encoding,) the server model is better. It's four 2.0 GHz cores (which can turbo up to 2.9 GHz on one core, 2.8 GHz on two cores, or 2.6 GHz on all four cores) as opposed to the desktop model's two 2.7 Ghz cores (which can turbo up to 3.4 GHz on one core, or 3.2 Ghz on two cores.)
I would be interested to see how thermal-constrained both models are. (AKA: How often they each turbo up.) If the server model is less thermally constrained, then it might turbo up to being faster on two cores than the desktop model (if the discrete GPU causes the whole thing to have thermal issues, it's possible.)
I would love to see some comprehensive light gaming benchmarks comparing the discrete graphics desktop model to the server model. (Which, if you upgrade both of their storage options to 750 GB spinning + 256 GB SSD, cost exactly the same, with the only difference being dual core 2.5GHz+discrete vs. quad-core 2.0GHz+integrated.)