Gadgetory

hey there my name's Gary Sims from Android Authority now two of the most popular mobile processor design companies at the moment are called Combe and Apple now Qualcomm's current leading processor is the snapdragon 8 to 1 and apple's of course is the a10 fusion but how do they compare with each other or they like when you put them side-by-side well if you're brave enough follow me and let's find out now before we get started with this comparison I just want to say two things first of all we are just talking about chips inside of a mobile phone there's enough problems in this world there's Wars that there's diseases there's terrorism ok let's not just blow this out of proportion and the second thing I'd like to say is that actually apple and Qualcomm aren't too forthcoming about the technology that's in their mobile processors so we have the snapdragon 801 and we have the apple a10 fusion well what are they like when you put them side-by-side what are their feature lists well let's have a look well first off it's worth mention that both chips are quad-core arm v8 compatible processors now Qualcomm and Apple both have architectural licenses from arm and that basically means they have the right to design their own arm compatible chips using their own designers using their own technology as long as it is compatible with arms instruction set and architecture now they've both got quad cores and they both use what's called heterogeneous multi-processing that's why I'm a polite Alling fusion what it basically means the two cores are kind of high performance they run at higher clock speed and they use more power and there are two cores which use less power on a lower clock speed and are more power efficient and the idea is that they juggle between the two depending on the different workload however there is one big difference Qualcomm can use all four calls simultaneously home from what we understand the Apple chip can actually only use two cores at one time and it switches between the two powerful calls and the two less powerful core now in terms of the GPU we have the Adreno 530 in qualcomm 8 to 1 and that's Qualcomm's greatest and latest GPU and kokum will be designing GPUs for several years now and we find them in their different series of chips as we've been going over the years now up until now Apple has used the power VR GPU from imagination technologies however this is the first time that Apple are actually shipping their own GPU now talking of GPU is also interesting to note that Apple supports OpenGL ES 3.0 and its own metal API whereas the Qualcomm Snapdragon supports OpenGL ES 3.0 and the Vulcan API and there is other api's so Apple have got less API support but then again they control the whole phone these chips are only made to go inside of iPhones and maybe inside of iPads whereas Qualcomm sell their chips so just about everybody that will buy them including Samsung including other smartphone makers including set-top box makers including tablet makers including media player makers they have produced a general specific will cater for lots lots of different use cases whereas Apple are producing one that only supports their use case and of course that's a difference in the business model between these two companies and one last thing worth mentioning the Qualcomm Snapdragon 801 actually has quick charge 3.0 support built into the processor so every phone that built that has the 8 to 1 in it has capable of having quick charge technology built into the phone every tablet the same whereas Apple moment doesn't have any quick tile charge technology and it's certainly not found inside of the SOC and in fact talking of things not included in the SOC another thing that's interesting is that the Qualcomm chip includes the LT modem as the modem used by your 4G and 3G in 2d networks built into the SOC where's Apple don't build the modem into their SOC the a-10 doesn't have a modem built into it but it does use modems on external chips and internally enough three out of four Apple phones use Qualcomm modem so Qualcomm actually are being used inside of Apple chips for their modem technology now my testing for performance comes in two categories first of all I'm using several of the off-the-shelf benchmark apps you find that work on both iOS and on Android so everyone will go and to - we've got bit Geekbench and we've got Basemark OS - now my second set of benchmarks are ones that I've written myself now writing benchmarks work across two platforms is quite of a challenge so the first set of benchmarks I use use the Lua programming language because there are SDK is available for both Android and iOS and I've actually got some C programs that I wrote and I've used those in a previous video that I've done when I was comparing the difference between C speed and Java speed on Android and I've taken the same code and I've compiled it on Android I've compiled it on iOS to see what the difference performance is so that's kind of a look now to do these performance tests I've got a Google pixel smartphone and I've got an iPhone 7 now as a caveat of course there may be faster Snapdragon 8 to 1 devices out there I chosen the picture because it's probably one of the most important phones of this year and of course I know that there is a difference maybe between the iPhone 7 and the iPhone 7 plus for example the iPhone 7 plus has got a bigger screen got a greater resolution screen to the GPU will have to work harder so you know I know these are just two phones and maybe we could use different phones to get different results however those are the two that I've got and they do give us a good indication of the underlying performance of the SOC s so our first test is with antutu and as we can see the iPhone 7 beats the pixel not by a huge amount but by a significant amount now if you go over to the androids comm website and find the article goes wrong with this video you'll see how I've analyzed the individual figures that come from the antutu results to see how it got these particular numbers and in particular it's worth noting according to an - to the pixel has a better GPU than the iPhone 7 however the rest of the test the iPhone 7 is actually faster and our next test of course is Geekbench and Geekbench has two types of tests one is a single core test which basically runs on one core in the phone - doesn't matter if there are two calls or four calls or eight calls it will give you the same result class of per core test and weak see that the Apple a 10 fusion is significantly faster than the snapdragon 8 to 1 when it comes to the single core test and in the multi core test basically all four cores are used simultaneously and again we can see that even in that multi core test the iPhone 7 came out tops and then really to verify the antutu and Geekbench are telling us the truth we use a Basemark OS 2 here which again shows a fairly significant win for the Apple a 10 fusion hats off to Apple that's an excellent result and definitely the numbers speak for themselves that was a results that we've got from these popular benchmark apps but now I'd like to use my own benchmarks to see what numbers we get out now my first test basically does a whole bunch of math processing it does some prime number finding it does some hashes it does some table sorting and basically it runs through the tests and it tells you how long it takes for each test to run now the time is in milliseconds and we can see that actually the Apple produced its result making it through all that tests in about half the time of the Snapdragon 8 to 1 so according to my first benchmark the other benchmark by an to 2 and geek bates seemed to be coming true the a-10 certainly did very well during this test now for my second test I move things over slightly into the graphics world I have a simulation which kind of puts two drops of water every frame into a container uses a physics engine to run that all that calculations and some floating crates that float around the water there's more water being added in and it basically accounts to see how many droplets of water can go in in a 90 second period now so the theoretical maximum score is ten thousand eight hundred so let's have see how these two phones did well the scores are almost identical the Apple a10 fusion in the iPhone 7 is ten thousand two hundred and two droplets of water in a 90 second time frame and the Samsung sorry the Snapdragon 801 with the Google picks or managed ten thousand won 178 so they're almost even so it seems as I bumped up the complexity here to a the whole 2d physical simulation with all those particles of water going around that actually both phones performed very very well and for my third set of tests we're now using these C programs I wrote both on iOS and on Android there are three different tests the first one produces some sha-1 hashes for a block of data does that lots and lots of times and then sees how long it takes the second one tries to calculate the first million prime numbers using trial by division and the third one is race it's a complicated maths function that does multiplies and divides as floating-point operations to see which function how long it takes to run that function so let have a look at the results now in this particular case lower is better and the Snapdragon 8 to 1 makes you complete my tests in nought point 6 5 4 seconds compared to naught point eight six four seconds for the iPhone 7 we see the same thing again with the prime numbers half the speed a half a time taken on the Snapdragon twice as long taken on the iPhone 7 and again for the math function again the same theme naught point 1 5 seconds compared to null point three oh three seconds now this is a bit of a mystery because up until now all the benchmarks we've seen have told us that the Apple a10 is significantly faster than the Snapdragon 8 to 1 in the Google pixel and yet when I run these C tests we find that actually it is the Snapdragon 8 to 1 which is twice as fast as the Apple a10 now I don't know why that is I just know that I compile these two C programs on the respective development tools that you have for each platform now there could be some problems for example maybe the optimization inside of the iOS app is less than the optimization on Android that's a possibility maybe there are the tests were too short for example maybe the tests ran on the lower cause in the iPhone and the schedule didn't have a chance to Swit them over to the high power cores because they only last as you've seen half a second or something like that so maybe the scheduler didn't get around to moving over to the high power calls or maybe there's some other kind of optimization that goes on in the software level that tells the hardware do things differently and that didn't kick in when I was running my tests so I don't know why but according to these tests of mine the snapdragon 8 to 1 is significantly faster than the a-10 now without further research I really can't say any more than that so it's all very well having the performance but what about the power how is the power usage of these two different mobile process as well as I've said I'm using the pixel and I'm using the iPhone now they've basically got different battery sizes in them there's different screen sizes in fact the pixel has twice the resolution in terms of pixels than the iPhone 7 so there's more power being used by the display so this is quite difficult to actually do so what I did and I this is really the best I could do right now is I turn both phones on turn them down to their minimum brightness level and left them for an hour sitting on the home screen of the respective operating systems and just saw how much battery was used and then when I ran the tests I actually - took away subtracted that power usage when it was idle hoping that will take away what the screen used so we can get maybe a closer look at what the system on chip is using so then what I did was I ran two tests one is I use the 3d game Epic Citadel to test the 3d capabilities of these two processes and on the second test I played a video to see what the usage was like so according to my calculations for doing the 3d gaming the pixel use 415 milliamp hours worth of battery whereas the iPhone 7 used 319 million pounds of battery so it might seem at first glance that the iPhone 7 is more power efficient than the Google pixel how we have to remember the Google picks is pushing around twice as many pixels per frame when it's running Epic Citadel and it's actually doing that without using twice as much power so although the pixel is using more power is actually doing more work and for the second test running the video we find actually both devices used are basically the same amount of power so even though the pixel is powering a higher resolution display the video processors in both phones work very very well and so now this year there's been millions of phones that will use the snapdragon 8 to 1 and the Apple a10 fusion I won't declare which is best because clearly in terms of performance the a-10 seems to have the lead but there are other factors involved like cost for example like API support like modem support like fast charging that you find inside of the snapdragon 8 to 1 what's the GPU like in the 8 to one compared to the GPU like in the a-10 i'll let the numbers speak for themselves you decide which one you prefer now I really will look forward to reading your comment but can I just again ask for a level of maturity we're talking about chips and a mobile phone we're not talking about anything else the world's got enough problems that it is without people getting angry with each other about which mobile phone they use if you want to get the technology if you want to admire what Qualcomm and Apple have both made then let's just talk about it in the comments below well my name is Gary Sims from Andrea thority I hope you enjoyed this video if you did please do give it a thumbs up also don't forget to subscribe to and Roth or --'tis YouTube channel don't forget to download our Android app because that will give you all of our latest news directly on your mobile phone and of course last but not least don't forget to go to Andrew authority calm because we are your source for all things Android Oh