Apple A12 benchmarks

[ksec]

https://m.weibo.cn/status/4236380060065313

~25% improvement in both Single thread and Multiple thread benchmarks.

Given the uniform improvement in both, this lead me to believe A12 may likely only be ~25% higher clock speed ( Max Clock 3Ghz ) 7nm A11. We don't know if there are any changes in other parts of SoC such as GPU.

I have long wonder when will Apple stop moving from node to node and IPC improvement YoY.

 

[FIVR]

Wow that's comparable single thread to an overclocked 8700k, isn't it?


So Apple is matching intel's 5ghz CPUs at greater than 100W with a 4Watt iPad chip.

 

[Mopetar]

So Apple is matching intel's 5ghz CPUs at greater than 100W with a 4Watt iPad chip.

Isn’t there an old saying about things that sound too good to be true?

I know Apple has an impressive team of chip designers, but they’re not so good as to cheat physics.

 

[FIVR]

Isn’t there an old saying about things that sound too good to be true?

I know Apple has an impressive team of chip designers, but they’re not so good as to cheat physics.

Maybe Moore's law is only dead for intel. Maybe their incompetence in advancing their process has resulted in an exponential decline in their efficiency in comparison to their competitors.


I don't see anything in the laws of physics that says "intel knows what they are doing".

 

[IntelUser2000]

The i3 8100 3.6 gets 5258 in single thread. Intel systems are noticeably faster on Android/Linux.

The 4.5W 7Y75 gets 4663 in single thread. If Intel was able to keep to their original schedule and Icelake was out late last year, we'd likely have seen Icelake Y chips go well above 5K.

 

[jpiniero]

Yeah you need to look at the individual tests and not just the overall score.

 

[FIVR]

The i3 8100 3.6 gets 5258 in single thread. Intel systems are noticeably faster on Android/Linux.

The 4.5W 7Y75 gets 4663 in single thread. If Intel was able to keep to their original schedule and Icelake was out late last year, we'd likely have seen Icelake Y chips go well above 5K.

Those chips produce an order of magnitude more heat than the chips used in iPads and iPhones. If you stuck a Core M 7Y75 in an iPad chassis is would have trouble maintaining a clock speed over 1Ghz due to thermal constraints.


These latest tests make it obvious that Apple will be dropping intel from MacBooks ASAP.

 

[IntelUser2000]

Those chips produce an order of magnitude more heat than the chips used in iPads and iPhones. If you stuck a Core M 7Y75 in an iPad chassis is would have trouble maintaining a clock speed over 1Ghz due to thermal constraints.

Order of magnitude means 10x. Why do people keep misusing that phrase?

Anand's own tests show the SoC power running at 3W running benchmarks for the SD845 and Exynos 9810 phone SoCs. Apple's Tablet SoCs are estimated to have a TDP of 5W, which is perfectly reasonable.

Intel for some reason can't get their Core platform idle power down to Atom/ARM level, but that's beside the point. We've been on the Skylake core for 3.5 years for one.

 

[FIVR]

Order of magnitude means 10x. Why do people keep misusing that phrase?

Anand's own tests show the SoC power running at 3W running benchmarks for the SD845 and Exynos 9810 phone SoCs. Apple's Tablet SoCs are estimated to have a TDP of 5W, which is perfectly reasonable.

Intel for some reason can't get their Core platform idle power down to Atom/ARM level, but that's beside the point. We've been on the Skylake core for 3.5 years for one.

It might not be 10x but it's at least 3-5x more heat produced by these Core M chips than the iPad. I own both (core M macbook and iPad pro) and the iPad is both faster and runs cool, which is very unlike the macbook. The macbook can become intolerably hot from 30 minutes of light microsoft excel work.


Intel has figured out how to get these chips to turbo up to 5Ghz for momentary benchmarks but when they run actual applications it turns into a disaster. Go buy a 12" macbook and you will quickly find out why "fanless" and "intel" should never have been combined in the same product.

 

[IntelUser2000]

I have a Viliv S5 UMPC that uses the Atom "Silverthorne", which is a 45nm chip with 2W TDP and gets quite warm.

I also had a Dell Venue 8 Pro that uses Atom "Bay Trail", and using the same TDP runs quiet at half the thickness, and double the battery life at 2/3rd the battery capacity!

Intel engineers had to make a huge leap in power management to make that possible, primarily because they wanted to make it suitable for Smartphones. It took them 4 years from the original Atom to achieve that.

For some reason, they can't do that with Core. Perhaps its because they want to keep the platforms separate. Perhaps its because they refuse to integrate the chipset since they want to keep the flexibility of having one chip scale from 4.5W to 100W.

Atom demonstrated it is possible. By lowering idle platform power, you can further lower heat generated in light usage.

 

[french toast]

Don't those 4.5w intel Y products have a base clock of 1.1ghz or something? ..it is not comparable really.
To make it fair you would have to take a 4.5w 10nm+ icelake on a bespoke socket to get a true comparison.
My bet is that 4.5w icelake would beat A12 on burst, but lose on sustained workloads, power efficiency, graphics.

 

[Abwx]

Order of magnitude means 10x. Why do people keep misusing that phrase?

.

Because they are not aware that s it s powers of ten that are used as basis, and consquently that half an order of magnitude is 3.16x.

 

[IntelUser2000]

My bet is that 4.5w icelake would beat A12 on burst, but lose on sustained workloads, power efficiency, graphics.

Yea, whatever management Apple is under, makes the teams thrive. Being the darling of the tech world also means top engineers naturally gravitate towards them.

Even in areas without Moore's Law, where gains are great and comes pretty easily, there are differences because individual's strengths are different. That is true of a team too.

 

[SAAA]

Maybe Moore's law is only dead for intel. Maybe their incompetence in advancing their process has resulted in an exponential decline in their efficiency in comparison to their competitors.


I don't see anything in the laws of physics that says "intel knows what they are doing".

You can be negative as you want about a cpu manufacturer but the x86 competitor isn't much better either. Seeing this mobile phone chip beating all existing Ryzen skus doesn't sound any alarm bell?

To put into perspective:

Intel Core i5-8400
2.8 GHz (6 cores) (4GHz turbo)
4878

AMD Ryzen 7 2700X
3.7 GHz (8 cores) (4.3GHz turbo)
4867

iPhone 8
Apple A11 Bionic @ 2.4 GHz
4218

A12 @ 3GHz (or whatever combination of IPC and clockspeed required for that result)
5200

I come to two conclusions, mutually exclusive:

-geekbench is crap and these scores mean nothing in the real world, like an Ax series desktop chip wouldn't run games and apps nearly as fast as the numbers suggest.

-scores are somehow valid thus all the genius CPU architects must have gone to Apple and are working on fantastic tech that unfortunately will never power our PCs.

Maybe common people happily buying 500-1000$ phones every two years but not even glancing at desktops until they die, proceeding then to replace those with 400$ trash class hard disk based notebooks has any correlation?

Somehow disheartening if true as I don't care a bit about a platform that has no compatibility at all to 99% of the programs I own and have used over the years, not to talk about all the money spent.

 

[FIVR]

Because they are not aware that s it s powers of ten that are used as basis, and consquently that half an order of magnitude is 3.16x.

I was aware it was a power of 10. Intel 8700k @ 5Ghz uses 100 Watts. I doubt this A12 uses more than 10 Watts.


That is a factor of 10, exactly.


you can also simply buy the products and see that the Macbook produces easily 5x as much heat as the iPad despite preforming worse. It also has about 2x the amount of battery power yet gets about 1/2 the battery life so you can tell the heat is being used up as power (worse efficiency).

 

[frozentundra123456]

You can be negative as you want about a cpu manufacturer but the x86 competitor isn't much better either. Seeing this mobile phone chip beating all existing Ryzen skus doesn't sound any alarm bell?

To put into perspective:

Intel Core i5-8400
2.8 GHz (6 cores) (4GHz turbo)
4878

AMD Ryzen 7 2700X
3.7 GHz (8 cores) (4.3GHz turbo)
4867

iPhone 8
Apple A11 Bionic @ 2.4 GHz
4218

A12 @ 3GHz (or whatever combination of IPC and clockspeed required for that result)
5200

I come to two conclusions, mutually exclusive:

-geekbench is crap and these scores mean nothing in the real world, like an Ax series desktop chip wouldn't run games and apps nearly as fast as the numbers suggest.

-scores are somehow valid thus all the genius CPU architects must have gone to Apple and are working on fantastic tech that unfortunately will never power our PCs.

Maybe common people happily buying 500-1000$ phones every two years but not even glancing at desktops until they die, proceeding then to replace those with 400$ trash class hard disk based notebooks has any correlation?

Somehow disheartening if true as I don't care a bit about a platform that has no compatibility at all to 99% of the programs I own and have used over the years, not to talk about all the money spent.

Glad you brought some perspective to this thread instead of just the intel bashing. I was just getting ready to post a question of whether AMD would score any better. Apparently the answer, as I expected, is no.

 

[dark zero]

The i3 8100 3.6 gets 5258 in single thread. Intel systems are noticeably faster on Android/Linux.

The 4.5W 7Y75 gets 4663 in single thread. If Intel was able to keep to their original schedule and Icelake was out late last year, we'd likely have seen Icelake Y chips go well above 5K.

On Linux Intel are faster but on Android nope... they have issues. That's why people recommend to emulate it with mid tier/high tier processors. Heck, even AMD suffers on Android. That's happens when an environment made for ARM is working on an x86 one.

 

[FIVR]

-geekbench is crap and these scores mean nothing in the real world, like an Ax series desktop chip wouldn't run games and apps nearly as fast as the numbers suggest.

-scores are somehow valid thus all the genius CPU architects must have gone to Apple and are working on fantastic tech that unfortunately will never power our PCs.

You can make all the excuses you want but in the end if the test was "stupid" or had no relation to performance intel would be able to game it and score just as well. Unless you are suggesting that Apple can "game" these test but intel is simply "too honest" to do the same.


The simple explanation is that something about intel's architecture is inherently less efficient than apple's.

 

[dark zero]

Actually is Apple which made their own uARCH so strong that they can handle themselves an OS with the similar features than Mac OS.

And if this is for phones, on Tablets or laptops will be FAR better.

Maybe the transition from x86 to ARM is totally real and Apple managed to show us.

 

[frozentundra123456]

You can make all the excuses you want but in the end if the test was "stupid" or had no relation to performance intel would be able to game it and score just as well. Unless you are suggesting that Apple can "game" these test but intel is simply "too honest" to do the same.


The simple explanation is that something about intel's architecture is inherently less efficient than apple's.

Apparently, since Ryzen scores very similarly to intel, as shown in the post above, they have the same "flaw". Really it is an apples to oranges comparison. Maybe we should ask to see benchmarks of the A12 running BF1 multiplayer.

 

[Space Tyrant]

Note that Mac OS scores very, very high in single thread on GB4. Different OS's have different overhead. Linux seems to have a small advantage over Windows in that benchmark, and Mac OS has a *big* advantage over everthing.

The A12 score may be reflecting some of that affect -- depending on the OS involved.

Edit: I may have to retract that... all of a sudden I'm seeing 9800 Windows on 8700K ST results!
Edit2: It looks like some of the high scores are odd in other ways: version 4.1 or 4.2.0, lower MT score than ST, 3 cores active, etc. Too much gaming going on in that BM?

 

[Thala]

You can be negative as you want about a cpu manufacturer but the x86 competitor isn't much better either. Seeing this mobile phone chip beating all existing Ryzen skus doesn't sound any alarm bell?

To put into perspective:

Intel Core i5-8400
2.8 GHz (6 cores) (4GHz turbo)
4878

AMD Ryzen 7 2700X
3.7 GHz (8 cores) (4.3GHz turbo)
4867

iPhone 8
Apple A11 Bionic @ 2.4 GHz
4218

A12 @ 3GHz (or whatever combination of IPC and clockspeed required for that result)
5200

I come to two conclusions, mutually exclusive:

-geekbench is crap and these scores mean nothing in the real world, like an Ax series desktop chip wouldn't run games and apps nearly as fast as the numbers suggest.

-scores are somehow valid thus all the genius CPU architects must have gone to Apple and are working on fantastic tech that unfortunately will never power our PCs.

Maybe common people happily buying 500-1000$ phones every two years but not even glancing at desktops until they die, proceeding then to replace those with 400$ trash class hard disk based notebooks has any correlation?

Somehow disheartening if true as I don't care a bit about a platform that has no compatibility at all to 99% of the programs I own and have used over the years, not to talk about all the money spent.

All the tests i have done as far with using Visual Studio 2017(MSVC) and compiling Win32 desktop apps for both ARM64 and X64 giving indication to the fact, that Geekbench is pretty reliable. However i am not blaming engineering talent.
My hypothesis is that x86/x64 will you only get that far independent how many engineering talent you throw at it. Its just an antiquated architecture, which relies on variable instruction length, microcode caches and predecode stages and less registers than any competing RISC architecture. No-one with a sane mind would come up with such an instruction set architecture today.

 

[Thala]

Note that Mac OS scores very, very high in single thread on GB4. Different OS's have different overhead. Linux seems to have a small advantage over Windows in that benchmark, and Mac OS has a *big* advantage over everthing.
The A12 score may be reflecting some of that affect -- depending on the OS involved.

You have to understand that the OS is irrelevant for Geekbench. Only the compiler has impact. Of course compiler choice correlates with the OS but the fallacy is to equate performance differences to the OS.
What we do know is, that gcc produces faster code than msvc, therefore we are seeing higher scores under Linux than under Windows.

 

[moinmoin]

Geekbench should use GCC (or LLVM) as compiler under all OSes to take compiler related variance out of the equation.

 

[FIVR]

Apparently, since Ryzen scores very similarly to intel, as shown in the post above, they have the same "flaw". Really it is an apples to oranges comparison. Maybe we should ask to see benchmarks of the A12 running BF1 multiplayer.

Did I ever say anything about Ryzen being any better?


The difference is that intel continues to arrogantly claim to have a process lead when all evidence suggest they are (at best) equal to the competition.


It is also obvious that Apple should be able to out-design AMD given that Apple's quarterly service revenue is equal to AMD's yearly revenue. It is far more embarrassing for a 200B company like intel to fail to compete so drastically.