Genza, 20 Dec 2017Samsung smartphones always get high scores on the benchmarks but the actual performance never ... moreYou're very wrong. Samsung's benchmark scores are usually the worst.
See SGS5 (QSD 801) vs OnePlus One.
See SGS7 (QSD 820) vs OnePlus 3.
See SGS8 (QSD 835) vs OnePlus 5.
Samsung doesn't overclock/clock the frequency as high as others. This gives the illusion that the Samsung devices should, for instance, perform like 15% slower than other devices. And in fact, they do. It's because while the processor is slower, they use faster storage, but then the software is slower, so things cancel out/get neutralised.
If you want to see what the QSD 845 can really do wait for Oppo to release the OnePlus 6 in May 2018. If they don't stuff the software, it should offer the best performance.... at least until the next Google Pixel comes out (the modding community works wonders).
Now you'll be able to post that picture of the crap that you eat 0.0001 seconds faster. Facepalm
AnonD-708613, 20 Dec 2017try shoot 4k60fps video using your smartphone. oh, you cant. see you in 2019 if you're luckyWell, the 845 supports 4k@60fps so.
Anonymous, 20 Dec 2017Maybe apple put a bigass cpu that never gets used fully, would over heat.. Besides their multi... moreNot exactly. Let's imagine that apple DID NOT use the RtS (Race to Sleep) Method those phones would have WAY worse battery life. (just to understand HOW crappy that would be. with ALWAYS ON monsoon cores running constantly. in GSMarena's test the endurance would be just 2 HOURS! (For the iPhone 8+).
I am an Android expert btw. but as a developer. I need to know my enemy too. (AKA iOS)
So. with RtS on. you get a decent battery life. and impressive preformance when your
program needs it. So. basically. the cores know how to scale their power to the app that needs it.
(Whatsapp for example doesn't need ALL the power of the core. in fact. when running a conditional test in a typical iPhone 8. the OS took it to run using the MISTRAL core! while when creating a song using Garageband. the OS took it to a Monsoon core. but kept power at about 42% load.
since there are a whole lot lof processes running background. it's much harder to explain why if so they are slow in comparation to real life preformance with an android phone)
as for the scores.
Let's do math. shall we?
I've preformed Geekbench in two modes just to understand what is the prowess of each core.
the Monsoon core is as expected a 4273 points. (using my iPhone as source.)
the MISTRAL core is about 2210 points.
let's round it to make it easier. shall we? 4300 MNS, and 2200 MST (MNS= Monsoon, MST= Mistral)
you'd think it'll be like 2xMNS(100%)+4xMST(100%)= Multicore.
which is 4300x2+2200x4=8600+8800=17400 points.
in fact. a multicore consumes only 50-75% of the power of the core.
so it'll be more like 3000+3000+1000+1000+1000+1000= 10000
a good example for that would be people working together. 2 are muscular, 4 are avarage. when they preform a task together. they DON'T use all their power.
Azaam, 20 Dec 2017Apple has double the scores but still it is slower to perform simple tasks compared to oneplus... moretry shoot 4k60fps video using your smartphone. oh, you cant. see you in 2019 if you're lucky
Azaam, 20 Dec 2017Apple has double the scores but still it is slower to perform simple tasks compared to oneplus... moreSimple. Apple's CPUs (Ever since the Ancient Apple A6) used a method called RtS. (Race to Sleep) which means they burst in full power to preform the task as fast as possible then get to minimum power in the same speed. to preserve battery life. (which is not that impressive. as we android lovers know) this is why in mundane tasks like fingerprint unlocking and messaging apps it seems slow. because the task executes so fast. (about 0.00006 seconds for a typical 100 MB app) that it seems slow.
the Oneplus series does NOT use the RtS method. as their cores are not as power hungry as apple's SoCs. so by atomic stopwatch count it takes about 0.00022 seconds to launch it. but since there are FOUR cores of power type. while apple has two powerful 4 efficient. speed is multiplied by 1.6 but for apple it's 1.2 which is why it seems FASTER on the android one but slower in iOS. but the differences are so negligible in these speeds it doesn't matter unless that ten thousandth of a second is crucial for you.
P.S. I'm a BSc in Software Engineering.
For an android that's amazing! I am still disappointed from Qualcomm that surely needs a push to reach the Monsoon core of A11. then again... the Monsoon core is HUMONGOUS compared to the Kryo 385.. so it's a tough task.
Anonymous, 20 Dec 2017It is greatly unimpressive. This single core score is taken straight from 2015.
Being 2.5 y... moreApple has double the scores but still it is slower to perform simple tasks compared to oneplus phones.
Check online videos comparing oneplus 5 vs iphone 7 or 8....
Opening simple apps, oneplus beats apple handsdown. (why didn't the 4500 geekbench score help?)
Oneplus fingerprint unlock is faster.
Apple is faster in opening video games and processing 4k video splitting kind of tasks,
Which are rare tasks compared to finger unlocking and opening dialer, messaging app kinda tasks.
Maybe apple put a bigass cpu that never gets used fully, would over heat.. Besides their multicore is crap compared to their single, around double and using 4 cores.
Android is exposed.... this benchmark of s9 is no match against iphone 8 and iphone X...
AnonD-708613, 20 Dec 2017this is spot on. are we going back to 2015 or what?May be crispy is right ..
Not need too much score ..
Thinking of the next iPhone's benchmark scores......
But like the article says, it's just a number.....
Thesis86, 20 Dec 2017The benchmark scores are impressive, given how this is just an incremental upgrade. It might b... moreIt is greatly unimpressive. This single core score is taken straight from 2015.
Being 2.5 years late is unimpressive, at this point Apple does double that.
As for processing power it is used everywhere, everybody "needs it", few "know" that they do.
A faster phone is faster. If you want to pay flagship money to get a phone that can barely match 2+ years old phones be my guest. As for me I'm not touching lazy Qualcomm, my next phone is either a Huawei (Kirin processor) or an Apple.
Qualcomm is in scramble mode. They either get their sh** together, or the android world gets left eons behind.
Btw Samsung could save this by using an Exynos SoC exclusively as they did in S6. But nooo, they keep injecting money in that useless company despite (probably) having a better processor. It makes no sense...
Genza, 20 Dec 2017Samsung smartphones always get high scores on the benchmarks but the actual performance never ... moreActually the real performance is much better from the benchmark numbers. Look at some speed battles on YouTube. s8 oreo version is neck to neck or even better from iPhone X with double the score.
Thesis86, 20 Dec 2017The benchmark scores are impressive, given how this is just an incremental upgrade. It might b... moreIf it helps them winning the clicking race, i'd say its worth the extra power. Fuckin android loses on every battle on youtube.
Benchmark-wise, a good upgrade if you coming from S6 or S7. Dont need to switch to this if you newly using your S8. Not for you. This S9 is just a more polished S8.
For me, the potential low power consumption/ possible supervisor battery efficiency is what I want to see. If the powers that be can create a system on chipset that provides top notch battery longevity, I might get a handset with a forthcoming second gen. 10 nm chipset next year. I wonder what 2019 handset options will be like (?).