PC Perspective’s Android to iPhone series explores the opinions, views and experiences of the site’s Editor in Chief, Ryan Shrout, as he moves from the Android smartphone ecosystem to the world of the iPhone and iOS. Having been entrenched in the Android smartphone market for 7+ years, the editorial series is less of a review of the new iPhone 6s as it is an exploration on how the current smartphone market compares to what each sides’ expectations are.

Full Story Listing:

iphonestorysmall.jpg

My iPhone experiment continues, running into the start of the third full week of only carrying and using the new iPhone 6s. Today I am going to focus a bit more on metrics that can be measured in graph form – and that means benchmarks and battery life results. But before I dive into those specifics I need to touch on some other areas.

The most surprising result of this experiment to me, even as I cross into day 17, is that I honestly don’t MISS anything from the previous ecosystem. I theorized at the beginning of this series that I would find applications or use cases that I had adopted with Android that would not be able to be matched on iOS without some significant sacrifices. That isn’t the case – anything that I want to do on the iPhone 6s, I can. Have I needed to find new apps for taking care of my alarms or to monitor my rewards card library? Yes, but the alternatives for iOS are at least as good and often times I find there are more (and often better) solutions. I think it is fair to assume that same feeling of equality would be prevalent for users going in other direction, iPhone to Android, but I can’t be sure without another move back to Android sometime in the future. It may come to that.

sleepapp.jpg

My previous alarm app was replaced with Sleep Cycle

In my Day 3 post I mentioned my worry about the lack of Quick Charging support. Well I don’t know why Apple doesn’t talk it up more but the charging rate for the iPhone 6s and iPhone 6s Plus is impressive, and even more so when you pair them with the higher amperage charger that ships with iPads. Though purely non-scientific thus far, my through the day testing showed that I was able to charge the iPhone 6s Plus to 82% (from being dead after a battery test) in the span of 1.5 hours while the OnePlus 2 was only at 35%. I realize the battery on the OnePlus 2 is larger, but based purely on how much use time you get for your charging time wait, the iPhones appear to be just as fast as any Android phone I have used.

Photo taking with the iPhones 6s still impresses me – more so with the speed than the quality. Image quality is fantastic, and we’ll do more analytical testing in the near future, but while attending events over weekend including a Bengals football game (5-0!) and a wedding, the startup process for the camera was snappy and the shutter speed never felt slow. I never thought “Damn, I missed the shot I wanted” and that’s a feeling I’ve had many times over the last several years of phone use.

emmaline.jpg

You don't want to miss photos like this!

There were a couple of annoyances that cropped up, including what I think is a decrease in accuracy of the fingerprint reader on the home button. In the last 4 days I have had more bouncing “try again” notices on the phone than in the entirety of use before that. It’s possible that the button has additional oils from my hands on it or maybe that I am getting lazier about placement of my fingers on the Touch ID, but it’s hard to tell.

Continue reading day 17 of my Android to iPhone editorial!!

In previous posts I have talked about the feeling of performance and interactivity being exceedingly high from a user experience.  But how do the iPhone 6s and 6s Plus with the Apple A9 SoC perform in relation to other processors in the mobile market? I ran a handful of quick benchmarks on the new iPhones, the OnePlus 2 with the Qualcomm Snapdragon 810, the LG G4 with the SD 808 and an iPad Air 2 just for another point of reference.

geekbench-2.png

geekbench-3.png

Geekbench is a synthetic test that measures the peak performance throughput of the processor in a series of integer, floating point, encryption and other benchmarks. The Apple A9 is unique in its design, a custom built processor from Apple that utilizes the same ARM 64-bit architecture as the Snapdragon 810, but only integrates 2 cores. These aren't off the shelf parts from ARM but rather fully custom design processors from engineers at Apple. The SD 810 is an 8-core processor while the SD 808 is a 6-core design. As a result you would expect Apple to have better single threaded performance with the A9 but fall a bit behind in the multi-threaded specific tests that can take better advantage of the additional cores. That is KIND OF what we see, but the Apple A9 is just a fantastic SoC either way – it dominates in the single threaded results but is capable of keeping pace with even the Snapdragon 810 in floating point and overall scores.

octane.png

sunspider.png

Both Sunspider and Google Octane are Javascript based benchmarks that attempt to evaluate some levels of real-world browser performance on mobile devices. We used Chrome on the Android phones and Safari on the iPhones to keep browser and OS-specific optimizations in place.

In both of these benchmarks the iPhone results are significant jumps ahead of the Android devices; it’s not really that close. I have had several people in the industry try to dissuade me from putting any emphasis on these tests but they are reasonable data points for a well-rounded discussion on SoC performance. Is the iPhone 6s 5-7x faster than the OnePlus 2? Absolutely not, but there is clearly something amiss about the Android results. Seeing an Android phone with a higher end processor run slower than one with lesser hardware (in both Octane and Sunspider) points us toward issues surrounding OS versions and update adoption.

3dmark-1.png

3dmark-2.png

3dmark-3.png

Measuring 3D graphics performance is a bit more direct and 3DMark shows a significant advantage for the iPhone 6s and iPhone 6s Plus when compared to these flagship Android devices. Based on the graphics score the Apple A9’s GPU is about 30% faster than the one found in Qualcomm’s Snapdragon 810, the company’s current flagship SoC. And though the gap lessens when we look at the Physics score, that utilizes the CPU cores rather than the GPU, the A9 still takes the lead.

gfxbenchnative-1.png

GFXBench has two modes it runs in during its testing process – offscreen and native. Native renders at the resolution of the screen of the phone while the offscreen result renders at 1080p on all devices. Clearly the differences change how you compare the scores: offscreen is what will allow you to make direct SoC to SoC comparisons while native results show you overall device performance. In this first result you can see that the iPhone 6s results in a score of 3466, 45% higher than the 2379 that the iPhone 6s Plus shows, despite using the same SoC and GPU. The reason is that the screen resolution on the 6s Plus is higher and thus the native result is rendered at that higher resolution, resulting in more work. The LG G4 has a 2560×1440 resolution screen, higher than any other phone tested here, and thus its score is significantly lower.

Why would you care about the native results instead of the more apples to apples offscreen scores below? Well, some games on these devices do in fact render at the screen resolution rather than a single common resolution and we have seen instances where performance is affected by higher resolution screens. Other apps render to a fixed resolution (like 1080p) and then the scale to the screen resolution (lower or higher) and avoid that concern.

gfxbench-1.png

gfxbench-2.png

In this case the scores between the iPhone 6s and iPhone 6s Plus are nearly the same, as both render at the same resolution with essentially the same SoC. Clearly the GPU on the Apple A9 is incredibly fast when compared to the other flagship devices on the market. Qualcomm has the Snapdragon 820 available soon and I’ll be curious how much of this story might shift with its release.

As I said, this is just a quick look at performance benchmarks on the new iPhone 6s. When the time comes for a full review we’ll have more devices in the comparisons in addition to more benchmarks covering different workloads. For my purposes here I wanted to provide enough data to back up my continued claims of perceived performance – the Apple A9 just FEELS fast, and the numbers prove that out.

You’ll see more updates from me in the near future concerning the iPhone 6s including an update of battery life and power efficiency when using our custom PC Perspective Wi-Fi browsing test.