Apple's A14 Chip Rumored To Become First Arm-Based Mobile Processor To Exceed 3GHz (macrumors.com) 34
Apple's A14 processor that's expected to debut this fall in Apple's iPhone 12 models is rumored to have a frequency reaching 3.1GHz. "This would be 400MHz higher than Apple's current A13 Bionic chip with a frequency of 2.7GHz," reports MacRumors. From the report: At such a frequency, the chip's Geekbench 5 running points have surged. The report mentions that the A14's single-core performance shows a score of 1658 (up 25% from the A13), and a multi-core score of 4612 points (up 33% from the A13). The extra processing power will be helpful in running simultaneous workflows, navigating through apps, and more. Apple chipmaker TSMC is expected to ramp up production of Apple's 5nm-based A14 chipsets in as early as April of this year. Also, according to 9to5Mac, Apple is reportedly planning to launch a new 5.5-inch entry-level iPhone with a solid state home button, Touch ID, and support for Apple Pay's Express Transit feature.
The 5.5-inch iPhone is expected to be priced $100 more than the 4.7-inch model that will start at $399, according to Apple analyst Ming-Chi Kuo. It's slated to launch at the end of March, but it's possible the launch will be delayed due to the coronavirus.
The 5.5-inch iPhone is expected to be priced $100 more than the 4.7-inch model that will start at $399, according to Apple analyst Ming-Chi Kuo. It's slated to launch at the end of March, but it's possible the launch will be delayed due to the coronavirus.
The chips are (Score:2)
Re: (Score:3)
Re: (Score:2)
If it'll cost you an arm and a leg at least you won't run faster.
I certainly hope so (Score:3)
No freakin' Face ID? I'll buy one.
Re: (Score:2)
What's wrong with FaceID?
Re: (Score:2)
Re:I certainly hope so (Score:4, Funny)
Re: (Score:2)
Re: I certainly hope so (Score:2)
Re: (Score:2)
Agreed. I don't have the TouchID turned on on my iPhone 7. I don't even have a passcode. I hate all of that crap.
I was thinking really hard about going back to a flip phone that can act as a wifi hotspot, but I will look at the low-end iPhones if they make them.
Apple extends its ARM (Score:2)
Gives Intel the finger.
Apple's ARM vs Intel's x86 (Score:3)
Apple's A14 was designed (power consumption, heat dissipation, etc) for small portable devices (iPhones).
Now imagine a similar design especially for laptops. More cores, etc. I'm pretty sure even the A14 could run circles around an Intel CPU of around the same power requirements as the A14.
When are ARM Macs coming? Will it be 2020, 2021, 2022? Knowing Apple and the fact that they like to control the hardware as much as possible and that Intel is more or less in the same position as IBM was more than a decade ago (PowerBook G5, anyone?), it would not surprise me if Apple dropped its contract with Intel.
Re: (Score:3)
Can't be far away now. Server-side has finally reached a point of comparable capability, vis Amazon's (Graviton2 [anandtech.com]), Marvell's (nee Cavium's) (ThunderX3 [anandtech.com]) and Ampere's (Altra [anandtech.com]) have all been released or announced, and all hold their own against contemporary cloudy Xeon kit. Impressively, Graviton2 (at least) seems to keep up against AVX512 on a single-threaded basis, mostly, according to the article linked. Proper server-grade memory interfaces make an enormous difference to real performance, compared to
The C/++ programming language kills it (Score:2)
If developers used a remotely modern approach like Java or .Net then the architecture is irrelevant as the JIT compiles the byte code at runtime. But people use C/++. An archaic system that compiles for the target architecture. And of course, does not normally utilize new instructions because they might not be available on all targets.
It is possible to decompile and then compile for a new architecture. The DEC Alpha did that long ago. But it is difficult to do well and not done.
So applications are lock
Re: (Score:2)
You should probably restrict expressing your ignorant views to websites where there's no chance that someone who knows anything might encounter them.
Re: The C/++ programming language kills it (Score:1)
Re: (Score:2, Insightful)
They are going full Pentium 4 on this thing, super high clock speed but in practice more cores is going to be of greater use to you.
If they want to go desktop/laptop they need to add a lot more PCI-e lanes and improve the memory controller to be more flexible. It's possible but not easy for a first generation version that doesn't suck. As with most Apple products it's probably best to avoid the first couple of iterations until they get it right.
Re: Apple's ARM vs Intel's x86 (Score:1)
Re: (Score:2)
No, Inte
Really... (Score:2)
Re: (Score:2)
What? They also "navigate through apps" and run "simultaneous workflows", things the new processor will be helpful for! Did you not read the summary?
Remember the good old days when you only ran sequential "workflows" on your phone? A simpler time...
Geekbench is bullshit (Score:3)
Geekbench is utter bullshit.
It seems to have iPhones outpacing modern workstations with over 10 times the power draw. Funny thing is if you write code and test it on a fast machine with fast, wide RAM and tons of thermal headroom it always seems to run faster than a small, thermally constrained chip with narrow memory interfaces and fewer cores.
Unless you're geekbench.
Re: (Score:2)
And yet every time someone whines about Geekbench they never, ever provide an alternative with better comparisons. In any case in artificial testing (like on a testbench with proper cooling and not stuck in a phone case) the existing A13 can do almost 4W/core. That's 8W for an iPhone, 16W for an iPad that could right into an ultraportable, 32W for a hypothetical 8-core laptop chip and 64W for a 16-core desktop chip. If I'm cherry picking the competition AMD for example sells the 12 core Ryzen 9 3900 @ 65W T
Re: (Score:3)
As an example, the crypto score alone makes the iPad pro 12 inch seem competitive with a desktop processor.
That's because that bad boy can apparently do AES on every core without any bottlenecks. Which is cool. But hardly a use case that anyone cares about.
That nearly doubles its multicore Geekbench score.
Re: (Score:2)
That's the real problem with Geekbench, and to a lesser extend many of these other ben
Re: Geekbench is bullshit (Score:1)
Re: (Score:2)
Who says there is one, and is the lack of one a reason to to point out the limitations of Geekbench? Perhaps people should be aware the limitations of tools.
I was not aware watts per core was a measure of performance. Are you claiming that the A13 is really a desktop processor if you simply repackage it?
Re: (Score:2)
And yet every time someone whines about Geekbench they never, ever provide an alternative with better comparisons.
Well, when you've got an oddball, locked down OS it's pretty hard to run an open test suite like Phoronix. That doesn't make Geekbench any more realistic though.
In any case in artificial testing (like on a testbench with proper cooling and not stuck in a phone case) the existing A13 can do almost 4W/core.
First I've seen those number, but OK sure... is that 4W for for each of the two big cores,
Thermal throttling (Score:2)
Nice, a CPU for an iPad on 3.1 Ghz. Will it behave like the thermal throttling i9 macbooks? I can see the poor thing getting very hot fast.