Antutu: iPhone 6s performance thrashes high-spec Androids Huawei Mate 8, Samsung Note 5 & Nexus 6p
Performance benchmark vendor Antutu presented iPhone 6s as the best performing smartphone of the year, despite being compared against high end Androids with more RAM and faster CPU clocks driving four times as many cores.
Source: Antutu
The company added Apple to its annual Performance Ranking Report for the first time this year, following the cross platform release of its Antutu 6.0 benchmark in 2015.
iPhone 6s, powered by Apple's custom dual core A9 chip with 2GB of RAM, debuted with a score 41 percent higher than the next fastest phone, Huawei's Mate 8, a phone that uses Huawei's internally designed octa-core Kirin 950 with 3 or 4GB of RAM (Antutu didn't specify which model it tested).
Apple's flagship score was 59 percent faster than Samsung's best ranking model, the Galaxy Note 5, a fablet powered by a higher-clocked Samsung Exynos 7420 Octa chip with 4GB of RAM.
iPhone 6s also scored 69 percent better than the Google-branded Nexus 6p, which uses a Qualcomm Snapdragon 810 octa-core chip and 3GB or RAM.
Since delivering its first generation 64-bit A7 powering iPhone 5s in 2013, Apple has remained ahead of more experienced mobile chipmakers in both performance and energy efficiency. Qualcomm, once a differentiated leader in mobile chips, struggled to deliver its first 64-bit chip last year, contributing to problems for Android licensees using its chips, including Xiaomi.
Qualcomm's missteps have also forced smartphone makers to develop their own chips, as Samsung, LG and Huawei have, albeit based largely upon generic designs drafted by ARM.
However, iOS also continues to be more efficient than Android in delivering performance from available hardware resources, and doing so with less RAM. Less installed system RAM helps contribute to longer battery life.
It also results in a device that's less expensive to build, contributing to overall profitability. Those profits are driving Apple to invest in advanced custom processor design, even as its rivals continue to largely use off the shelf core designs published by ARM in an effort to keep prices low.
Rather than being a "halo" high-end model, Apple's fastest iPhone models make up the vast majority of its sales, contributing to an iPhone Average Selling Price of $670. Other vendors of high end Androids primarily sell low end or mid grade budget smartphones, resulting in an ASP for Android closer to $250.
That perpetuates a cycle of profitably for Apple, while driving premium buyers to iOS. Apple has benefitted from a regular influx of switchers, with its chief executive Tim Cook noting in September that switchers from Android had reached 30 percent of its sales.
Antutu's benchmarks are so closely watched by enthusiasts that in previous years, Samsung and other companies have deliberately worked to cheat the company's benchmarks to achieve a phony ranking, simply for bragging rights.
In 2013, Antutu changed its benchmarks specifically to block cheating, after Samsung along with Asus, HTC and LG, were repeatedly caught fudging their performance rankings.
"With the exception of Apple and [Google subsiderary] Motorola," AnandTech observed, "literally every single OEM we've worked with ships (or has shipped) at least one device" that has cheated in benchmarks.
Source: Antutu
The company added Apple to its annual Performance Ranking Report for the first time this year, following the cross platform release of its Antutu 6.0 benchmark in 2015.
iPhone 6s, powered by Apple's custom dual core A9 chip with 2GB of RAM, debuted with a score 41 percent higher than the next fastest phone, Huawei's Mate 8, a phone that uses Huawei's internally designed octa-core Kirin 950 with 3 or 4GB of RAM (Antutu didn't specify which model it tested).
Apple's flagship score was 59 percent faster than Samsung's best ranking model, the Galaxy Note 5, a fablet powered by a higher-clocked Samsung Exynos 7420 Octa chip with 4GB of RAM.
iPhone 6s also scored 69 percent better than the Google-branded Nexus 6p, which uses a Qualcomm Snapdragon 810 octa-core chip and 3GB or RAM.
Apple advancing its 64-bit lead through profitability
Since delivering its first generation 64-bit A7 powering iPhone 5s in 2013, Apple has remained ahead of more experienced mobile chipmakers in both performance and energy efficiency. Qualcomm, once a differentiated leader in mobile chips, struggled to deliver its first 64-bit chip last year, contributing to problems for Android licensees using its chips, including Xiaomi.
Qualcomm's missteps have also forced smartphone makers to develop their own chips, as Samsung, LG and Huawei have, albeit based largely upon generic designs drafted by ARM.
However, iOS also continues to be more efficient than Android in delivering performance from available hardware resources, and doing so with less RAM. Less installed system RAM helps contribute to longer battery life.
It also results in a device that's less expensive to build, contributing to overall profitability. Those profits are driving Apple to invest in advanced custom processor design, even as its rivals continue to largely use off the shelf core designs published by ARM in an effort to keep prices low.
Rather than being a "halo" high-end model, Apple's fastest iPhone models make up the vast majority of its sales, contributing to an iPhone Average Selling Price of $670. Other vendors of high end Androids primarily sell low end or mid grade budget smartphones, resulting in an ASP for Android closer to $250.
That perpetuates a cycle of profitably for Apple, while driving premium buyers to iOS. Apple has benefitted from a regular influx of switchers, with its chief executive Tim Cook noting in September that switchers from Android had reached 30 percent of its sales.
Cheating didn't win
Antutu's benchmarks are so closely watched by enthusiasts that in previous years, Samsung and other companies have deliberately worked to cheat the company's benchmarks to achieve a phony ranking, simply for bragging rights.
In 2013, Antutu changed its benchmarks specifically to block cheating, after Samsung along with Asus, HTC and LG, were repeatedly caught fudging their performance rankings.
"With the exception of Apple and [Google subsiderary] Motorola," AnandTech observed, "literally every single OEM we've worked with ships (or has shipped) at least one device" that has cheated in benchmarks.
Comments
However, to all the Android users who love to throw Antutu results around like they're gospel, well this has got to sting. As in Japanese Giant Hornet sting.
Why do you hate them (I don't know much about them myself)? They seem to be just a quasi-analsys group pushing benchmark data.
No. According to Samsung themselves, the 8890 will be "up to 30% faster than the 7420 and 10% more power efficient".
Sorry, but your bogus charts of made-up numbers don't mean squat. I'll take Samsung's official statements about the 8890 instead.
The 8890 is also still only a 3-wide design (Apple A7-9 are all 6 wide). They're going to have to crank the clock way up AND add cores to even begin to catch the A9.
Oh, and the Geekbench numbers on your chart are also wrong. Primate Labs lists the 7420 at 1340/4512. This chart shows them at 1485/5129. Already inflated. Interesting, though, that your chart lists the "correct" Primate Labs figures for the Apple A9. Curious why they chose to do that. Never mind, not curious at all. With the rampant cheating on the Android side it only makes sense they used some one-off test figures for the 7420, instead of the thousands of aggregated tests that Primate Labs based their numbers on.
They are unreliable. And there have been lots of questions about how much weight they give certain results over others before they release their "final" score. Back when the iPhone 6S came out it scored 59000 while the 6 score 46000. Now they are showing 80000 and 132000? Not only are the scores completely different (changed their methodology again), but the RATIOS are different.
Now what happens when people are trying to compare results and they look at results before and after Antutu did their "update". Makes things much more confusing. Further, are the current scores more accurate? Or were the previous scores more accurate?
Stop lying and stop peddling fantasies.
Why are you here? You presence here makes no sense, unless your fracking trolling.
Considering how it's been in the last few years.... This fall is when irremediably fall behind in performance. They're already 1 year behind in their high end phones.
I'm an iPhone user and I think benchmarks matter. They give a good idea as to the capabilities of a company (like Apples processor engineers). They're also a great way to measure progress from generation to generation.
To Android users the benchmarks only matter if one of their preferred devices gets a better score (usually in one cherry-picked result). When the iPhone is better (which is most of the time) then benchmarks suddenly don't count. Or they're not directly comparable across different OS versions. Or they're not comparable across different processor architectures (Intel guys or Apple haters who refuse to believe the A9X is challenging Intel for performance in mobile processors). Or Geekbench is optimized for iOS vs Android.
The Android side is so full of disastrous UI tweaks by vendors that benchmarking is such a frivolous thing. Apple doesn't even need a faster CPU to deliver superior performance to almost any Android phone. I'm actually certain at this point the iOS animations are probably the bottleneck and has been for a processor generation or two.