LG G6 vs iPhone 7 vs Galaxy S7 vs Google Pixel vs Huawei P10: performance benchmarks comparison

LG claims that using the Snapdragon 821 with some optimizations as the system chip for the LG G6 is a 'brilliant move', despite the fact that the upcoming Samsung Galaxy S8, for instance, is expected to ship with a chip that will be newer and more powerful, the Snapdragon 835...
This is a discussion for a news article. To read the whole news, click here


15. ebilcake

Posts: 1230; Member since: Jul 16, 2016

None of these tests measure sustainable or long term performance so it's irrelevant. Does not help that iPhone 7 and Huawei P10 run lower resolutions than the other devices either, really wish sites would stop wasting our time with these benchmarks.

20. mikehunta727 unregistered

Not sure what site's you visit to see wasteful benchmarks but yeah this speciific test by PA is quite rubbish, Anandtech and Arstechnica and others do great accurate reliable benchmarking

21. ebilcake

Posts: 1230; Member since: Jul 16, 2016

They use the same benchmarks, although they do cover more of the tests.

22. mikehunta727 unregistered

Yes they use the same benchmarks but their methodology is much better and more scientific and accurate then PA'. More granular detail on the results

51. ebilcake

Posts: 1230; Member since: Jul 16, 2016

Lots of Android sites and reviewers continually post results from these benchmarks, they usually completely ignore the offscreen results and long term performance. I'm not even sure Anandtech bothered with long tern performance.

58. mikehunta727 unregistered

Lol of course Anandtech tests long term perf, they have many times

88. ebilcake

Posts: 1230; Member since: Jul 16, 2016

I just looked back at their Galaxy S7 review and no, they never ran these tests. They did their own tests looping the benchmarks .

24. jeroome86

Posts: 2314; Member since: Apr 12, 2012

Dang. That midrange iPhone beating all the flagships. That's pretty good.

43. manbearpig

Posts: 10; Member since: Jan 19, 2017

iPhone 7 is a midrange? LOL

61. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

Yea it is. I dont know any high-end devices with 1080 or 750p displays...high-end devices have 1440p, mid-ranged have 1080p and the low end have 800p or less. So yeah, the jokes on you. BUB! You probably think a car with 4 cylinders is a high-end car...even if its has Benz on the outside. Last I checked a 4 cylinder is at the bottom, even if it is a luxury brand.

81. drahcir02

Posts: 102; Member since: Mar 03, 2016

Xperia devices with 1080p

89. ebilcake

Posts: 1230; Member since: Jul 16, 2016

You're not fooling anyone with this nonsense.

95. trojan_horse

Posts: 5868; Member since: May 06, 2016

He ain't fooling anyone, by the way.

26. darkkjedii

Posts: 31063; Member since: Feb 05, 2011

Damn, the iPhone on point.

28. marorun

Posts: 5029; Member since: Mar 30, 2015

They obviously not yet put the final version of GPU drivers. So thats low GPU scoore can be fix with a software patch anyway.

48. ebilcake

Posts: 1230; Member since: Jul 16, 2016

Nope, you really don't have a clue do you? LG G6 =1440x2880 So guess what happens when you take a Snapdragon 821 and run on on-screen graphics test, yes, the results suck balls. Drivers won't be fixing that, ever.

62. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

Its not gonna suck. What games have you played that sucked on 1440p?

90. ebilcake

Posts: 1230; Member since: Jul 16, 2016

I don't use a 1440p device, I use OnePlus 3 which doesn't throttle GPU performance. However a lot of the 1440p devices do throttle quite heavily like the Galaxy S7, thankfully I avoided having to experience that this year or have manage resolution. This is a problem you don't seem too understand, the benchmark results you see are meaningless, when you're playing a graphics heavy game for up-to an hour or more performance on the Galaxy S7 and other devices will tank (probably after as little as 10 minutes), there are plenty of games where you will notice this over time. You can solve it by changing resolution to but you shouldn't have to do this. Of course increasing to 1440*2880 is only going to make matters worse, higher resolutions equals lower framerates but also more stress on the SoC which means more throttling.

52. cnour

Posts: 2305; Member since: Sep 11, 2014

As usual, iPhone is in a very high place comparing to the others.

57. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

There aren't any GPU demanding games on mobile. If there are, someone please list which ones so I can download and try them.

59. mikehunta727 unregistered

Vainglory, CSR Racing 2 come to mind. Enable highest res and highest graphical options in app settings

66. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

Installing Vainglory. It basically reminds me of z Kingdom as far as Graphics. Being a PC gamer, I play games that I know are GPU intensive. I am talking a game like SimCity 5 for example, which has 1000's of moving sprites on the screen, that is also handling extensive mathematical calculations in the background and then has huge 3D graphics. A game which I have pushed at fill 4K on my 4K display and via HDMI to my Samsung 4K TV. Yes the game does lag a bit at full 4k. Runs like a dream at FHD. But even when I lower the resolution to achieve 60FPS, other than the whole screen moves faster, everything else remains exactly the same. I don't think mobile games are to intensive for the mobile GPU. What I think though is, some devs take more advantage of some GPU options vs others. In general Android devices are more powerful than iOS devices based on specs. Yet devs try to make games better on iOS and then make them lame on Android. The devs who do make them equally as good on Android, you see their is hardly any difference visually or in gameplay.

77. mikehunta727 unregistered

Simcity is not a GPU intensive game, it requires a much more beefier CPU then it does GPU for the game, Simcity is actually one of the most CPU intense games to play on a PC, and they are mostly still poorly multi threaded and don't use more then 2 cores on the game engine logic If you are connecting your 4K TV via HDMI, then the max refresh rate of your TV is 30 hz and can't show more FPS past 30 hz, which would feel pretty laggy to play on a 30 hz TV. I play on a 4K monitor via Displayport, allows for 60 hz at 4K and I have Freesync enabled also to get rid of screen tearing in games And well yes the game's out aren't pushing our SoCs to 100% because then battery life would tank fast and also sustained performance would drop like a fly on many devices so they find that balance in graphical quality and sustainability for most devices In general, Android devices are actually not more powerful then Apple's A series counterparts, and the A10 The reason why many iOS games also generally look better on iOS vs Android (more graphical effects/etc) is because of developers using Apple metal on A series SoC's to get noticeably more performance/watt out of the SoC's, allows for better graphical rendering efficiency and more stuff can be done on the screen without performance penalities of past graphical API's. On android, most games are still using OpenGL, which is not as low level as Metal is to iOS games, which means worse performance and worse battery life at the moment until Vulkan really takes off the ground running on Android. At the moment, Vulkan has very barely taken off on Android so..

80. mikehunta727 unregistered

Unless you have a HDMI 2.0 cable then 4K/60 FPS is possible, otherwise you'd be using displayport *

93. ebilcake

Posts: 1230; Member since: Jul 16, 2016

What the hell has Simcity 5 got to do with mobile gaming? Simcity 5 is from 2013.... :-/ Vainglory is well optimized by default and will run on almost anything as it scales down the resolution if you don't enable HQ mode.

63. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

I play CSR Racing...me an my kid. The graphics on my Note 7 are just as good as they are on any other phone. It doesn't lag or have bad graphics. I also play Asphalt 8, Need for Speed an more. The Need for Speed No Limits takes advanatge of the extra benefits of the Note 7 with HDR10. The game looks better than it does on my iPhone 6S Plus and Galaxy S7 Edge

92. ebilcake

Posts: 1230; Member since: Jul 16, 2016

Doesn't lag? That's debatable given results in speedtests the Samsung devices always lose. Samsung devices do throttle quite a lot though so your long term performance is probably pretty rubbish. Although you have the advantage that nobody is likely to ever confirm that on the Note 7 for fear of the obvious. Why are you still using one? Being a fanboy is one thing but continuing to use recalled device is just stupid.

65. TechieXP1969

Posts: 14967; Member since: Sep 25, 2013

I don't know why PA loves Antutu. It proves nothing. There are to many variables to consider to ever consider a score. Fact, the only phone sold on the market where the phone is 100% identical on any carrier is the iPhone. Yet when you use Antutu, its own scores are not consistent. Then when you test Android, lets use Smasung as prome example, even phone in the USA even with the same hardware is not identical, do to each carrier adding their own bloat. The only naked option is if you buy a Exons model from Samsung with no blaot. Even then you get very different score. For fun I ran 5 tests back to back on my Note 7. I forgot to screenshot the lowest score which was 89,XXX. I assume since the Note 7 and the S7 Edge have the same hardware, they should bench within reason of being in the same ballpark. Yet the 4 tests I ran after the first all gave very inconsistent numbers.https://flic.kr/s/aHskV6tvV3 - From 107506~142459. I bet if I used Package Disabler an disable every possible piece of bloat, I'd likely reach or suppose the iPhone 7's numbers. SO using benchmarks doesn't tell ANYTHING. The numbers are simply not consistent, are based on way to many variable and they don't even give a rough estimate. Considering each phone is different based on carrier/ region, what apps come preinstalled and what apps the end user will add, the benchmark shows nothing. Its not different than spitting off a roof 10 times and thinking your spit will land in the exact same spot. Benchmarks are as artificial as fake flowers. Anyone who would use a benchmark as some type of way to lay claim to how fast or slow a device is, is just some retard fan-boy who has nothing else to lay claim and uses an artificial inconclusive benchmark as some type of arsenal to argue with. "Oh look my device is faster because it has a higher benchmark"...lol - Said only by an idiot!

73. darkkjedii

Posts: 31063; Member since: Feb 05, 2011

79. mikehunta727 unregistered

"I bet if I used Package Disabler an disable every possible piece of bloat" Doesn't help benchmark scores, have done it in the past on my GS6 Edge Plus with over 120 things disabled with Package disabler pro, did not increase my benchmark scores in any meaningful way at all And antutu will pretty much never give the same exact score, even on two exact same pieces of hardware running the program, each will be off by a few hundred points determined by the benchmark's algorithims every time.
This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at https://www.parsintl.com/phonearena or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit https://www.parsintl.com/ for samples and additional information.