iPhone X Geekbench BLOWS any Android phone away

[TheReal1980]

It seems the new processor in the iPhone X is a real beast.

Single core : 4061

Multi core : 9969

 

The new Note 8 scores:

1815 and 6066.

 

How is it possible?

 

http://www.gsmarena.com/iphone_x_spotted_on_geekbench_iphone_7s_appears_on_antutu-amp-27224.php

 

[PCGuy_5960]

Meh, Geekbench is a really bad benchmark. 

[JediFragger]

They've worked out a way to combine all 6 cores (2 high performance and 4 efficiency ones) to get an ultra-performance mode. Pretty cool.

[Guest]

4 minutes ago, PCGuy_5960 said:

Meh, Geekbench is a really bad benchmark. 

For PC yes. 

For mobile? Maybe not. 

[Drak3]

Just now, deXxterlab97 said:

For mobile? Maybe not. 

No, Geekbench is an all around shit benchmark.

[Sauron]

10 minutes ago, TheReal1980 said:

How is it possible?

Geekbench is a bad benchmark.

3 minutes ago, deXxterlab97 said:

For PC yes. 

For mobile? Maybe not. 

For everything, it's just bad. It has no reliable correspondence with real life performance. And you know there's something wrong when the multicore score on a dual core cpu is more than twice the single core score.

[Aytex]

How is it possible? Create your own OS and optimize it around your custom arm chip and utilize all it's resources? 

[Bouzoo]

12 minutes ago, TheReal1980 said:

How is it possible?

Benchmarks can differ in what they test (which workload), and can be optimized for specific workloads/hardware, among other things. Same how we got info that Ipad Pro had a better CPU than i5s at the time, which is utter nonsense. You could probably optimize your benchmark so that Snapdragon 835 beats an i7. Wait for proper reviews, don't believe everything you read on the first.

Read more here.

 

Let's try and cherry pick some of them:

 

Benchmark 1:

 

Benchmark 2:

[Doobeedoo]

While Geekbench may not be famous I really wonder why they're the only one going with such core cluster design opposed to Android top chips. 

Less cores overall but focus more on pair of heavy performance ones and secondly on low power ones. More single threaded performance, still great multi and efficiency is there. 

[Sauron]

Just now, Aytex said:

How is it possible? Create your own OS and optimize it around your custom arm chip and utilize all it's resources? 

This should not affect synthetic benchmarks significantly. The benchmark should take care of grabbing all the resources it can get. All you may see is some slightly lower performance due to ram usage by the OS, but that's negated by the fact the iphone has a third of the ram everyone else has.

[MeDownYou]

iPhone x who are these guys Nvidia????

 

[Aytex]

1 minute ago, Sauron said:

This should not affect synthetic benchmarks significantly. The benchmark should take care of grabbing all the resources it can get. All you may see is some slightly lower performance due to ram usage by the OS, but that's negated by the fact the iphone has a third of the ram everyone else has.

Still is more powerful than anything Android has to offer. Qualcomm needs to die and mediatek needs to fuck off with their shitty healio SoC's

[mynameisjuan]

Or you could look at pixel and nexus devices and see how hardware has nothing to do with performance if optimized correctly. 

 

I mean I would still love to see how fast android could be on apple hardware

[Sauron]

Just now, Aytex said:

Still is more powerful than anything Android has to offer. Qualcomm needs to die and mediatek needs to fuck off with their shitty healio SoC's

Uhm says you... based on a botched benchmark. Besides considering it has an advantage of several months a slight performance edge is expected. Let's wait for some reliable numbers that actually have meaning before we jump to conclusions, shall we? Not that it really matters, 99% of customers will use it for selfies and facebook anyway and those have worked just fine for the last 3 years on anything but the crappiest 50$ pieces of junk.

[mynameisjuan]

7 minutes ago, Aytex said:

Create your own OS and optimize it around your custom arm chip and utilize all it's resources? 

No way! Who would of thought having tight control of hardware and software would lead to great performance and fast updates?

 

Like I dont think many people can grasp the challenge android has. Like with windows having to support a plethora of hardware indirectly.

[Sauron]

6 minutes ago, MeDownYou said:

iPhone x who are these guys Nvidia????

It's the same people who brought you mac OS X

[Aytex]

2 minutes ago, Sauron said:

Uhm says you... based on a botched benchmark. Besides considering it has an advantage of several months a slight performance edge is expected. Let's wait for some reliable numbers that actually have meaning before we jump to conclusions, shall we? Not that it really matters, 99% of customers will use it for selfies and facebook anyway and those have worked just fine for the last 3 years on anything but the crappiest 50$ pieces of junk.

Well yeah of course. I completely agree with the last part ? but we all know apples A chipset is far superior than Android's. There's not much to be superised by from this benchmark

[Drak3]

5 minutes ago, MeDownYou said:

iPhone x who are these guys Nvidia????

 

iPhone 10, for the 10th anniversary of the product line.

 

3 minutes ago, mynameisjuan said:

Or you could look at pixel and nexus devices and see how hardware has nothing to do with performance if optimized correctly. 

Both product lines you mention perform exactly as one should expect of their hardware. Niether are all that optimized.

[TacoSenpai]

Meanwhile,  I'll still be perfectly content with my S8+ for many years to come provided its still working.

[AlTech]

20 minutes ago, PCGuy_5960 said:

Meh, Geekbench is a really bad benchmark. 

Well, one could argue that it's good at showing how bad the app actually is .

[Sauron]

Just now, Aytex said:

we all know apples A chipset is far superior than Android's. 

Well... again, I don't know how you can state that without proper testing. Not to mention Android is not a manufacturer. In the past few years Apple has shown better single core performance and worse multi core performance, as you would expect given the core counts of their chips compared to their competitors. I wouldn't say one approach is inherently better than the other, it all depends on the software and since Android phones have had quad cores for years now every developer who makes cpu intensive apps has optimized them accordingly at this point. I do suspect apple sticks to dual cores simply because of backwards compatibility with apps that wouldn't use more than two cores anyway.

[AlTech]

Just now, Sauron said:

Well... again, I don't know how you can state that without proper testing. Not to mention Android is not a manufacturer. In the past few years Apple has shown better single core performance and worse multi core performance, as you would expect given the core counts of their chips compared to their competitors. I wouldn't say one approach is inherently better than the other, it all depends on the software and since Android phones have had quad cores for years now every developer who makes cpu intensive apps has optimized them accordingly at this point. I do suspect apple sticks to dual cores simply because of backwards compatibility with apps that wouldn't use more than two cores anyway.

actually,

the A11 chip breaks from tradition.

 

There's a mixture of core counts.

High Performance : 2 cores delivering 30-40% faster performance compared to 2 previous gen High Performance Core

Low Power: 4 cores delivering 70% faster performance compared to 2 previous low power cores.

 

70% extra perf for 100% extra cores seems like some slightly bad scaling.

[mynameisjuan]

Just now, Drak3 said:

iPhone 10, for the 10th anniversary of the product line.

 

Both product lines you mention perform exactly as one should expect of their hardware. Niether are all that optimized.

Ummm or you could compare said benchmarks between them and said iphone that was released at the same time and do a performance test and even though the iphone smokes it in benchmarks, in real world use the iphone loses in response time.

[Sauron]

1 minute ago, AluminiumTech said:

actually,

the A11 chip breaks from tradition.

 

There's a mixture of core counts.

High Performance : 2 cores delivering 30-40% faster performance compared to 2 previous gen High Performance Core

Low Power: 4 cores delivering 70% faster performance compared to 2 previous low power cores.

 

70% extra perf for 100% extra cores seems like some slightly bad scaling.

Which is akin to qualcomm's designs, but the point is that in high performance mode it's still a dual core. The lower performance scaling on the low power cores might be due to lower clock speeds for better power efficiency, that would have to be tested.

[AlTech]

6 minutes ago, Sauron said:

Which is akin to qualcomm's designs, but the point is that in high performance mode it's still a dual core. The lower performance scaling on the low power cores might be due to lower clock speeds for better power efficiency, that would have to be tested.

Yes but Apple unveiled their new 2nd gen Performance Controller which is able to utilize all the CPU cores at once in some situations. So......

 

EDIT: Linked Verge video at time.