Benchmarks show Android Ice Cream Sandwich is fast, but Samsung’s chips are faster

galaxynexus

Now that the Samsung Galaxy Nexus is officially out, at least in Europe, it’s time to see how it performs compared to the competition. We can’t think of a better site to run benchmarks on this flagship device than Anandtech, because they’re so meticulous that it’s damn near scientific. Let’s start with the dual core 1.2 GHz ARM Cortex A9 application processor. Using the SunSpider JavaScript test, the Ice Cream Sandwich running Galaxy Nexus completes the benchmark in 1878 milliseconds, faster than the 2250 milliseconds required by the iPhone 4S running iOS 5. The same performance advantage is seen in BrowserMark, where the the Nexus scores 98,272 BrowserMarks (what the hell is a BrowserMark?) while the iPhone 5 manages to pull only 87,841. To put that into perspective, this writer used the same benchmark on his 2010 17 inch MacBook Pro with a 2.53 GHz Intel Core i5, 4 GB of RAM, running Google Chrome 15.0.874.121 on Mac OS X 10.6.8 and managed to score 493,650 BrowserMarks; my computer is 5x “better” than the third generation Google Phone. Could a faster clocked Apple A5 compete with ICS? Probably, which makes the A6 all the more exciting.

Now onto the graphics processor inside the Texas Instruments OMAP 4460 that powers the Galaxy Nexus. It’s a PowerVR SGX 540, the same GPU found in the Samsung Hummingbird processor that’s at the heart of the 18 month old Samsung Galaxy S. Using the GLBenchmark 2.1 Pro test, it’s 39% slower than the ARM Mali-400 GPU found within the Exynos 4210 that powers the Samsung Galaxy S II. Score that next to the PowerVR SGX 543MP2 inside the iPhone 4S, and it’s 91% slower! Should you be concerned by this? Well, yes and no. If you’re a gamer, then you might have some problems, not just because the GPU is slower, but because some games might not run on Ice Cream Sandwich since it’s so different compared to Gingerbread. As for general usage, everyone who has reviewed the Galaxy Nexus says it flies, that it’s buttery smooth, that it’s the closest Android has ever been to iPhone touch responsiveness.

At the end of the day though it needs to be said that benchmarks don’t matter. They really don’t. The iPhone 4S and Samsung Galaxy S are so different on so many levels that looking at a bar graph doesn’t mean diddly-squat. It’s tough to recommend one over the other since it really depends on what you want out a device.

  • androidymus

    Its a little funny that you post these benchmark tests and then write as a conclusion, that they dont matter at all. Why did you post it then in the first place?

    • http://www.intomobile.com/ Stefan Constantinescu

      For the people who do care. Maybe I can convince them not to care?

    • Anonymous

      Yeah that was kind of corny. It reads to me as if the writer noticed that his test would spark controversy. It would show that the Samsung chips are faster and that the iPhone 4s outdid the Samsung Galaxy S II overall. So to stay neutral, the Benchmarks become of no importance.

  • Jon Garrett

    benchmarks are nice but to the average consumer, they not only don’t care but they don’t even know what benchmarks are. this is true for both iOS and Android.

    Only we care. :-)

  • Khalid W

    Line 16 has a typo of “iPhone 5″. That’s still due next year my good friend ;).

Back to top ▴