On Apple’s A7 Processor | Hacker News Reader

On Apple’s A7 Processor

stevecheney.com

18 pointsMikeCapone19 comments

"whether the chip is dual or multicore, 32-bit or 64, you can bet Apple has it all tightly woven together"

I don't get it. Any device maker has multiple options in those four configurations. All mobile chip makers seek to optimize "performance per Watt".

What parameters are Apple going to tweak to make it "exactly fit" rather than suffer the heartbreak of a "generic design compromise"?

Reading over the article, it doesn't really say anything of technical content other than "Apple has complete control and can vertically integrate, applications have different needs per thread, if you can tightly couple hardware and software, it'll be fast." Computer architects have known that for decades by now.

Apple isn't custom designing the IPCores for the processor or graphics. I'm willing to bet that almost all of this chip is off the shelf stuff. What they can do is choose parameters like number of cores, cache sizes, CPU options (like you said, stuff that is available with existing suppliers) and also include additional units that they can license separately (e.g. earSmart)

The blog post mentions stuff like Frameworks and Apps. Unless I am very out of touch with SoC offerings these days, you don't get a MapKit.framework accelerator. You get things like H264 encoding/decoding and AES, or DSP functions which these frameworks can use. However, the faster performance you want with one function, the more locked in you get. An example of this is comparing encryption speeds on an Intel processor with AES-NI. AES is an order of magnitude faster than just about everything else. Everything else however, uses the same vector extension set that were designed specifically to speed up sub operations of common algorithms.

The other advantage they get by doing an in house chip is that they can choose the on board peripherals they want (possibly remove some chips that would normally be external), and the exact number of external busses they need for all the external peripherals, which can lead to a smaller, cheaper, more power efficient chip.

Unlike Android, Apple needs 900000+ third-party apps binary compatibility so they can't tightly weave anything, except perhaps extract a little more power savings from the kernel/CPU interface.

Only if you ignore the gigantic set of APIs that Apple also controls through which those binary apps get the vast majority of their work done.

Kind of hope they stay dual-core and keep improving the work done per clock. Right now, too little software will use four cores. They could add distinct low-power cores (e.g., so Siri can listen while the phone is sleeping) or other specialized stuff, but four identical cores just doesn't strike me as the best use of die area in a mobile device.

The question will be "can they beat Intel" once Intel gets its shit together and becomes serious about mobile. And the answer would be an unequivocal "no".

The burden of proof certainly is on Intel right now, not Apple.

But even if Intel was to make great mobile chips, Apple sells because of the whole package, not just the chip. What does Intel have on the mobile side, except the possibility of being one more supplier for semi-anonymous Android and Windows phones?

Intel has been doing the groundwork for using their chips with Android. If they deliver on Intel-quality chips, and they will, these will, due to process alone, blow anything ARM out of the water, both on power consumption and performance.

Thats a huge win for Android, and by extension it will be a huge win for Intel. It certainly seems like Intel is heading to Android for an entry into the market.

If Intel wins big -- they will be the supplier for Apple. Apple switched to Intel in the past, they might do it in the future if it is dominant enough.

Apple is Intel's #1 customer.

Apple has been clearly saying that they want as much control as possible over every aspect of their products. Lots of vertical integration.

Unless Intel does something amazing that can't be duplicated by others like they have done in the PC market, I don't see why Apple would give up its chips, which are the best in mobile and very customized for the company's needs.

In mobile, Intel is behind. On the PC side, they were way ahead. Yesterday's winners aren't always today's...

Intel is pivoting the entirety of their chip design/production infrastructure, it's a big ship to turn. I wouldn't expect anything serious from Intel until Q4 2014.

> it's a big ship to turn

Indeed. And they're taking their time turning it too. Atom was first announced in 2007 (according to Wikipedia), just 3 months after the first iphone was released. Look at Apple now, while Intel are still busy turning that ship...

What kind of margins can they expect, though? The Wintel duopoly was very good to them, but nothing similar exists in mobile (except maybe Apple chips + iOS). They can't be as dominant and successful, with great pricing power, as just another supplier...

Android might be a challenge as well if Samsung locks up their supply chain as tightly as Apple.

That being said, it'd be easy for them to put out iOS on atom if needed. I'd imagine they'd have gotten good at this with Marklar.

There's probably a good enough working relationship between Apple and Intel that they can probably figure out how to get a custom silicon design from Intel if push came to shove.

cache: http://webcache.googleusercontent.com/search?q=cache:stevech...

That being said, isn't the Facebook iOS app known for being a huge CPU hog? Every time I boot it off the multitasking tray, my iPhone (original 4) feels like a new phone...

The design they describe is how you do anything on iOS that's going to go out and talk to some web services, otherwise talk on the network, or the like. You do your UI stuff on the UI thread and you use additional background threads for the network stuff.

In any case, if you're not in the Facebook app, then iOS has frozen it and it's not getting any CPU time at all. Removing it from the multitasking tray does not stop it from consuming any CPU resources, because if Facebook isn't on the screen then it's not running.

The Facebook iOS shouldn't be running once it's off your screen. I really loved the iOS app actually. I went from S3 to iPhone 5 to HTC One. The Android app was fine when I was on the S3, but it seemed to have gotten worse while I was on iOS.

Battery usage on my HTC One has several times been double what my screen uses (very, very, very weird) even if I've swiped the app away from my recent apps after using it for maybe 10 minutes.

Either way, I've always been amazed at how terrible a company like Facebook is at designing a coherent mobile experience. Their Symbian, Windows Phonee, Android, and iOS apps are all different from each other and just hit (recent iOS is mostly fine) or miss (all the others). I think they have a different team for each platform.

Reply on news.ycombinator.com