Primate Labs Blog

Geekbench 2.2.7

2012-02-20T00:00:00-08:00

Geekbench 2.2.7 has been released and is available for download from the Geekbench website. Geekbench 2.2.7 features the following changes:

Fixed an unhandled exception on OS X 10.8 (Mountain Lion).
Fixed a bug that prevented "Check for Updates..." from working properly.
Fixed a bug that misidentified the iPhone 4 (CDMA).
Improved processor frequency detection on Windows.
Improved processor detection on Linux.

Geekbench 2.2.7 is a free update for all registered Geekbench 2 users.

Geekbench 2.2.6

2012-01-19T00:00:00-08:00

Geekbench 2.2.6 has been released and is available for download from the Geekbench website. Geekbench 2.2.6 features the following changes:

Fixed a hang that could occur when running benchmarks on Windows.
Fixed a bug that canceled benchmarks prematurely on Android.
Fixed a bug that misidentified some late model MacBook Pros.
Re-enabled multi-threaded tests on single-core systems.
Improved compatibility with Red Hat Enterprise Linux 5.
Improved upload error messages on Windows and Linux.
Improved handling of corrupted Geekbench files.

Geekbench 2.2.6 is a free update for all registered Geekbench 2 users.

Geekbench 2 for Android

2012-01-14T00:00:00-08:00

Happy 2012! To start off the new year I'm pleased to announce that Geekbench 2 for Android is now available on the Android Market.

Geekbench 2 for Android uses the same tests as other versions of Geekbench 2, meaning the results are comparable across platforms. Not only can you compare the performance of over 600 different Android devices, you can also compare the performance of Android phones against Apple iPhones, or Android tablets against Windows laptops.

Geekbench 2 for Android, like Geekbench 2 for other platforms, is also multi-core aware. This enables Geekbench to measure the performance of today's dual-core Android phones and tomorrow's quad-core Android tablets.

Finally, Geekbench 2 for Android works with the Geekbench Result Browser. You can upload, share, and compare your Android results with people all over the world.

Geekbench 2 for Android is available today in the Android Market for only $0.99.

Geekbench 2.2.3

2011-11-02T00:00:00-07:00

Geekbench 2.2.3 has been released and is available for download from the Geekbench website, and the iTunes App Store (it's currently under review for the Mac App Store). Geekbench 2.2.3 fixes most outstanding issues with Geekbench 2.2. A complete list of all changes can be found in the Geekbench release notes.

Geekbench 2.2.3 also removes support for Mac OS X 10.4. Users who need to run Geekbench on Mac OS X 10.4 can download Geekbench 2.2.0 or Geekbench 2.1.13.

Geekbench 2.2.3 is a free update for all registered Geekbench 2 users.

iPhone 4S Benchmarks

2011-10-11T00:00:00-07:00

While the iPhone 4S won't be released until Friday, Geekbench results for the iPhone 4S have started to appear on the Geekbench Result Browser. These initial benchmark results are quite promising; the iPhone 4S scores almost 70% higher than the iPhone 4. The iPhone 4S also compares favorably to both iPads; the score is less than 20% lower than the iPad 2, and almost 40% higher than the original iPad. See the following chart below to find out how the iPhone 4S performs compared against all of Apple's iOS devices.

These initial results also confirm that the iPhone 4S comes with 512MB.

iOS Performance

iPad 2	746
iPhone 4S	622
iPad	447
iPhone 4 (CDMA)	371
iPod touch (fourth generation)	370
iPhone 4 (GSM)	368
iPod touch (third generation)	274
iPhone 3GS	272
iPod touch (second generation)	185
iPhone 3G	135
iPhone	135
iPod touch	132

Geekbench 2.2

2011-08-31T00:00:00-07:00

I'm pleased to announce Geekbench 2.2 has been released and is available for download from the Primate Labs website. Geekbench 2.2 is a free update for all registered Geekbench 2 users.

Geekbench 2.2 contains a number of new features and bug fixes. While you can read about these features and fixes in the Geekbench release notes I'd like to highlight some of the new features I'm really excited about in this release.

The most noticeable change is that Geekbench is now one application instead of two or three separate applications. You can now run 32-bit and 64-bit benchmarks from the same application; it's no longer necessary to stop one version and start another version when you're analyzing your computer's performance.

Another change is noticeable after running the benchmarks; the result window has been redesigned. I worked with Matthew Skiles to come up with a fresh and vibrant look for the Geekbench results window.

The biggest new feature, though, is the new hardware stress tests (available under the "Benchmarks" menu). The stress tests load all of the available processors, cores, and threads to uncover any problems that might be lurking, undetected, in your computer.

The stress tests have been a huge help here at Primate Labs. I've used the stress tests to test new computers to ensure they're working properly, and I've also used them to track down an underperforming heat sink on an older workstation.

Finally, Geekbench 2.2 includes command-line tools on both Mac OS X and Windows. While most users will prefer the "traditional" graphical user interface, the command-line interface provides enterprise users a way to automate benchmarks.

You can find out about all the changes in Geekbench 2.2 from the Geekbench release notes. Geekbench 2.2 is the first major release since I went indie earlier this year, and I think it's the best release yet. I hope you think so, too.

July Mac Benchmarks

2011-07-30T00:00:00-07:00

There are enough results in the Geekbench Result Browser for last week's new Mac models that I can now update the Mac Benchmark charts. To highlight the performance of the new Mac mini and the new MacBook Air I've provided charts below that compare the new Macs against other Macs. This allows you to easily see the performance improvements with the new Macs.

The results below are based on 32-bit results from the Geekbench Result Browser (both 32-bit and 64-bit results are available on the Mac Benchmark page). Geekbench scores are calibrated using the original entry-level Power Mac G5 as a baseline with a score of 1,000 points. Higher scores are better, with double the score indicating double the performance.

Mac mini

I've compared the latest generation of Mac minis against the previous two generations of Mac Minis. Here are the results:

Mac mini (Mid 2011) Intel Core i7-2635QM 2.0 GHz (4 cores)	8586
Mac mini (Mid 2011) Intel Core i7-2620M 2.7 GHz (2 cores)	6977
Mac mini (Mid 2011) Intel Core i5-2520M 2.5 GHz (2 cores)	6465
Mac mini (Mid 2011) Intel Core i5-2415M 2.3 GHz (2 cores)	5842
Mac mini (Late 2009) Intel Core 2 Duo P8800 2.66 GHz (2 cores)	3697
Mac mini (Early 2010) Intel Core 2 Duo P8800 2.66 GHz (2 cores)	3627
Mac mini (Late 2009) Intel Core 2 Duo P8700 2.53 GHz (2 cores)	3526
Mac mini (Early 2010) Intel Core 2 Duo P8600 2.4 GHz (2 cores)	3310
Mac mini (Late 2009) Intel Core 2 Duo P7550 2.26 GHz (2 cores)	3204
Mac mini (Early 2009) Intel Core 2 Duo P8400 2.26 GHz (2 cores)	3102
Mac mini (Early 2009) Intel Core 2 Duo P7350 2.0 GHz (2 cores)	2818

The Sandy Bridge processors in the new Mac mini provides a tremendous boost in performance: the quad-core Mac mini server is 2.3x faster than the previous Mac mini while the dual-core Mac minis are 1.6x to 1.9x faster.

Another interesting thing (not shown in the chart above) is that the Mac mini server has roughly the same performance as the entry-level Mac Pro. While this will not hold true after the next Mac Pro update (which is overdue), it's impressive to see Apple's smallest computer provide comparable performance to Apple's largest computer!

If processor performance is your primary concern, and if you don't mind limited internal upgrade options, the Mac mini provides a tiny, power-efficient, and reasonably priced alternative to the Mac Pro.

MacBook Air

I've compared the latest generation of MacBook Airs against not only the previous generations of MacBook Airs, but also against the current generation of MacBook Pros. Here are the results:

MacBook Pro (17-inch Early 2011) Intel Core i7-2820QM 2.3 GHz (4 cores)	10370
MacBook Pro (15-inch Early 2011) Intel Core i7-2820QM 2.3 GHz (4 cores)	10304
MacBook Pro (15-inch Early 2011) Intel Core i7-2720QM 2.2 GHz (4 cores)	9980
MacBook Pro (17-inch Early 2011) Intel Core i7-2720QM 2.2 GHz (4 cores)	9963
MacBook Pro (15-inch Early 2011) Intel Core i7-2635QM 2.0 GHz (4 cores)	8775
MacBook Pro (13-inch Early 2011) Intel Core i7-2620M 2.7 GHz (2 cores)	6864
MacBook Pro (13-inch Early 2011) Intel Core i5-2415M 2.3 GHz (2 cores)	5910
MacBook Air (13-inch Mid 2011) Intel Core i7-2677M 1.8 GHz (2 cores)	5826
MacBook Air (11-inch Mid 2011) Intel Core i7-2677M 1.8 GHz (2 cores)	5796
MacBook Air (13-inch Mid 2011) Intel Core i5-2557M 1.7 GHz (2 cores)	5353
MacBook Air (11-inch Mid 2011) Intel Core i5-2467M 1.6 GHz (2 cores)	4557
MacBook Air (13-inch Late 2010) Intel Core 2 Duo L9600 2.13 GHz (2 cores)	3013
MacBook Air (13-inch Late 2010) Intel Core 2 Duo L9400 1.86 GHz (2 cores)	2678
MacBook Air (11-inch Late 2010) Intel Core 2 Duo U9600 1.6 GHz (2 cores)	2268
MacBook Air (11-inch Late 2010) Intel Core 2 Duo U9400 1.4 GHz (2 cores)	2021

Like the Mac mini, the new Sandy Bridge processors provide a tremendous performance boost to the MacBook Air lineup: the 13-inch MacBook Air is 1.9x faster than the previous 13-inch MacBook Air, while the 11-inch MacBook Air is an incredible 2.6x faster.

The Core i7-based MacBook Air is almost as fast as the Core i5-based MacBook Pro ( only 20% slower). While there has always been a (sometimes sizable) gap between the fastest MacBook Air and the slowest MacBook Pro, this update eradicates that gap. Now it's no longer a question of whether users want a light laptop or a fast laptop; it's a question of whether users want a light laptop or a heavy laptop.

Unfortunately, users who don't just want a fast laptop but the fastest laptop still have to pick the "heavy" 15- or 17-inch MacBook Pro.

What About Arrandale?

Looking at these results I'm surprised at how long Apple stuck with the Core 2 processor. Intel's Core i5 and Core i7 "Arrandale" processors have been out for for 18 months, and have models suitable for both the Mac mini and Macbook Air. Why did Apple wait for the release of the Core i5 and Core i7 "Sandy Bridge" processors in order to make the switch?

It turns out the answer lies with graphics chips. Mac OS X relies heavily on graphics chips, or GPUs, and needs a graphics chip that performs well. While the other Macs can use discrete GPUs, both the Mac mini and MacBook Air have space constraints that rule out discrete GPUs. This means Apple can only use integrated graphics in these models.

Thanks to Intel's lawsuits against NVIDIA, Apple couldn't use NVIDIA's integrated chipset graphics. The only option Apple had was Intel's integrated chipset graphics, which are known for their poor performance. So while the Arrandale processors would've provided a performance boost the integrated graphics would've provided a performance drop.

Sandy Bridge, however, ships with integrated graphics that are "good enough"; Apple can use Sandy Bridge to provide processor performance improvements while assuring decent graphics.

Final Thoughts

Overall I'm really impressed with the updates to the Mac mini and MacBook Air. Now that Apple has moved away from Intel's aging Core 2 architecture, users have great performance across the entire Mac lineup.

Benchmark Updates

2011-06-10T00:00:00-07:00

I updated a number of Geekbench benchmark charts today. The biggest update was to the PC benchmark chart, which was last updated in January 2009. Whoops. Unfortunately, a lot of infrastructure work needed to happen before I could update the PC benchmark chart. It was hard to schedule this work before I went indie. Now, though, I scheduled a big chunk of this week to improve the infrastructure, and now that it's done updates will happen more frequently.

I also updated the Mac benchmark chart to include the results for the latest iMac models. Finally, I updated the iMac benchmark chart from last month to include the 21.5-inch Core i7 iMac (the results weren't available when I first published the chart).

I hope you find the updated charts useful!

Going Indie

2011-05-15T00:00:00-07:00

Earlier this year my wife accepted a post-doctoral fellowship in Portland, Oregon. As part of the relocation we realized I'd need a job in Portland (staying in Waterloo while she was in Portland wasn't an option we were willing to consider). We realized that working on Primate Labs was actually an option; revenue has grown each year to the point where it started to look like a full-time endeavor.

Thus, I'm excited to announce that I'm "going indie". Thursday was my last day as a full-time Intel employee, and Monday will be my first day as a full-time Primate Labs employee. I'm also excited to announce that I'm (temporarily!) relocating to Portland, Oregon.

Now that I'm full-time, what will change? I'll be spending more time improving my products. I've got a lot of ideas for great new features and improvements for Geekbench and Wiinote and I'm looking forward to implementing them over the coming months. I'll also be able to provide faster responses to support requests.

As part of this transition I'd like to thank my customers for their patronage, feedback, and support. I'd also like to thank my wife, Deborah, for her patience and support over the four years of Primate Labs' existence. Working both a full-time job and a part-time business has meant we've had to make a lot of sacrifices. I'm hoping that by going indie I'll have more time for both my customers and for me and my family.

iMac Benchmarks (Early 2011)

2011-05-05T00:00:00-07:00

Earlier this week Apple released new iMacs which feature Intel's next-generation quad-core Sandy Bridge processors. To see how well the Sandy Bridge iMacs perform I've collected Geekbench 2 results from the Geekbench Result Browser for these new iMacs and compared them against the previous Wolfdale (Core 2 Duo) and Lynnfield (Core ix) iMacs. While results aren't available for all the Sandy Bridge iMac configurations (in particular I don't have results for the 21.5-inch Core i7 iMac) there are enough results here to examine the performance of the new iMacs. Update! Results are now available for all of the Sandy Bridge iMacs (including the 21.5-inch Core i7 iMac). I've updated the chart below with the results.

For people not familiar with Geekbench 2, higher scores are better. Also, Geekbench 2 only measures processor and memory performance; it won't measure the benefits of new video cards or storage devices.

Finally, if you're curious how your computer stacks up against the new iMacs you can download Geekbench and find out (it's a free download).

iMac Benchmarks

iMac (27-inch Mid 2011) Intel Core i7-2600 3.4 GHz (4 cores)	11648
iMac (21.5-inch Mid 2011) Intel Core i7-2600S 2.8 GHz (4 cores)	10248
iMac (27-inch Mid 2010) Intel Core i7 870 2.93 GHz (4 cores)	9123
iMac (27-inch Mid 2011) Intel Core i5-2400 3.1 GHz (4 cores)	8357
iMac (27-inch Late 2009) Intel Core i7 860 2.8 GHz (4 cores)	8335
iMac (21.5-inch Mid 2011) Intel Core i5-2500S 2.7 GHz (4 cores)	7890
iMac (27-inch Mid 2011) Intel Core i5-2500S 2.7 GHz (4 cores)	7844
iMac (21.5-inch Mid 2011) Intel Core i5-2400S 2.5 GHz (4 cores)	7241
iMac (27-inch Mid 2010) Intel Core i5 680 3.6 GHz (2 cores)	6934
iMac (21.5-inch Mid 2010) Intel Core i5 680 3.6 GHz (2 cores)	6828
iMac (27-inch Mid 2010) Intel Core i5 760 2.8 GHz (4 cores)	6726
iMac (27-inch Mid 2010) Intel Core i3 550 3.2 GHz (2 cores)	5922
iMac (21.5-inch Mid 2010) Intel Core i3 550 3.2 GHz (2 cores)	5854
iMac (21.5-inch Mid 2010) Intel Core i3 540 3.07 GHz (2 cores)	5681
iMac (Late 2009) Intel Core 2 Duo E8600 3.33 GHz (2 cores)	4647
iMac (Late 2009) Intel Core 2 Duo E7600 3.06 GHz (2 cores)	4221

While the improvements aren't as dramatic as with the Sandy Bridge MacBook Pros, there's a consistent 25% performance improvement between Sandy Bridge and Lynnfield at both the low-end and the high-end of the iMac model range (and over a 70% improvement between Sandy Bridge and Wolfdale). While this update may not tempt Lynnfield iMac owners into upgrading, it's certainly a compelling upgrade for Wolfdale iMac users.