I can see it. I just don't have much use for motion technology for something that is normally in my hands all the time (I love wings, but I don't eat them and use my phone that much
)
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
A7 Processor to Be 31% Faster and 64-Bit, iPhone 5S to Support Motion Tracking?
- Thread starter MacRumors
- Start date
- Sort by reaction score
I can see it. I just don't have much use for motion technology for something that is normally in my hands all the time (I love wings, but I don't eat them and use my phone that much
)
I'm not talking about things in terms of any OS in particular. From the perspective of an embedded systems architect I'm thinking in terms of the processor itself. Seriously, what capabilities or performance benefit does a larger memory space get you when all of your physical RAM and a good portion of your filesystem can already be mapped into a single process address space? It's not like most iOS devices have a ton of memory mapped I/O. The MMU works the way the MMU works, and what it can do really doesn't work any differently if you're talking about the Windows kernel or a UNIX/OS X/OS of the week kernel.
What you were describing earlier was memory isolation via process address space mapping. 32 bit processors do it now, and they're pretty good at it. Moving to 64-bit processors isn't going to help them do it any better. If I can cover all my physical address space with 4G of address range and I can remap each process to it's own range (with plenty of dead space to spare) what do I get?
Last edited:
Yes.
My point was more that there is a mistaken notion that more cores means it will work a certain amount better. A dual to quad will mean twice as fast, for example. What they don't realize is that it doesn't work like that. The quad core might not even be much faster, if any, if the software only uses dual cores. That is why I say people are obsessed with numbers.
I'm admitting that petty first world advertising has spoiled me into want 2x as fast.
I won't base anything on the new processor, I doubt I'll upgrade because it makes me feel too frivolous since I have a functioning iPhone.
----------
It's not a standard, just a feature I was expecting. And I doubt I'll upgrade because I have a working phone and it makes me feel unappreciative.
I like it! I am quite moved indeed.
And this might have a 2x as fast GPU for your number pleasure.
I believe Apple will certainly try to leapfrog ARM whenever possible and practical. They have the license to modify the architecture and spent hundreds of millions to build up a team dedicated for that purpose. Also Apple controls their own development tools and thus they can move ahead quicker (or slower) with the hardware progress as they see fit.
So far I haven't read anything about the GPU side of things and often times that's the more interesting part of Apple's decision with iOS hardware. Has there been any leaks regarding the GPU?
We've been hearing about that for a long time and there have been speculations on Rogue for a while now. However my question was whether there's anything of substance with some evidence or any information on the specifics on the GPU.
While OS X facilitates this, it is primarily a function of the processor and the x86-64 architecture. Not all 64-bit processors can run 32-bit code. ARM AArch64 (the 64-bit ARM architecture) instruction set is dramatically different from the 32-bit instruction set.
Check out this discussion to see notes from the implementors of the linux AArch64 port.
Apple's really good at transitions though. I wouldn't expect major problems with a 64-bit transition (fat binaries?). However, I'm seriously not sure what benefit they would have by going 64-bit right now.
Not sure that's a good thing.....just got an iPhone 5. It's battery life is pathetic, much worse than my iPhone 4....
We have no such thing here in Canada. Our terms are three years, with no early termination fee. We simply buy out the remaining cost of the subsidized phone.
When I went from iPhone 4 to 5, I paid $180, plus $200 to buyout my phone. I sold my phone for $360. Ultimately, I paid $20 to upgrade a two year old phone. If I originally had the iPhone 4S, I would have had to pay more to buyout, but recovered the cost with the higher resale value. So upgrading is essentially free. $20 to have a new phone for a year (until the next version is released)? Sounds fine to me.
Barring some ridiculously high cost for this new iPhone, upgrading after only a year is a no brainer. Sure, I am extending my contract for a year, but that means nothing. Here, there is no way to pay a lower monthly fee if I own the phone straight up, so the is no benefit to not having a contract.
As for the topic of this thread, I hope the new iPad also gets the A7 (or A7X).
And if security was painless, more people would do it.
Most people wouldn't buy insurance if it wasn't required in most activities.
Apple has made a 'no' decision to focus on portability and discrete 4, 8, 10, 11, 13, 15 inch sizing for mobile, assuming that those are the critical few for sweet spots. Yes, you'll get closer to 100% coverage if you fill in all the gaps, but that's SKU drift that Apple doesn't want to excurse into.
Now... a 'no password' security system (first fingerprint, then true voice/facial recognition (linking voice/face and movement) is painless security (except in below zero climates like MN, where a bellaclava and gloves may make it problematic ;-). And that is critical in moving to a new economic system that is both cashless, and signatureless:
Something you have (a device)
Something you are (your face, voice, fingerprint)
Something you know (the final, vs the first level of security... the secret).
For 50 years, passwords have been the first level vs the last (typically, security is 'where you are', 'who you are', 'what you know [secret phrase]'. We can get back to that, starting with fingerprint (requiring a warm, pulsing finger, not a picture of a fingerprint, but a 3D and multi-layer (surface, subsurface, and capillary)). Couple that with a few years of being able to get to voice prompting (Please say your name and passphrase), and then a 3d facial recognition (again, defeating paper pics), and we are on our way to a level of 'consumer security' that will greatly reduce the losses that 'ASCII passwords' now expose us to as an 'only form' of online security.
Aah I see. Well, as a simple consumer i don't care
I just want a responsive beautiful phone.none of this is data that supports fact. Reasons for switching phones can be vast and I would be willing to guess (note that I did not say it is fact) that the #1 driving force is price.
The 5S will have noticeably better battery life then the 5, more energy efficient processor that is also 31% faster, faster, more efficient GPU....slightly bigger battery, possibly they are going to use IGZO display in the 5S?
Probably no IGZO but i can defitnetly see the 5S getting a few more hours of batterry life in like everything
You're right, but on the other hand a newer chip running at a 31% higher clock speed can be assumed to be at least 31% faster in performance than the old one, as performance (of the chip itself, at least) tends to roughly scale with clock speed and the newer chip probably won't be slower per clock cycle).
That's the problem I have with the Tweet and this article. We do not know how many cores the chip has, what kind of improvements have/haven't been made to improve the per clock performance, the RAM type used, etc, etc. Without those information, the figure 31% is utterly useless but yet Macrumors and other internet sites took it and ran with the story as if it really showed the new chip will be "31% faster". It's really sad to see the utter lack of basic critical thinking and writing in tech blogs.
(or just as likely, they don't care as long as they get their clicks and ad revenue)still not better than the sanpdragon S4 pro (LAST YEAR's processor). believe me apple, color options will not save you.
Don't matter. ios will run better than any android device on those snapdragons. It's just too well optimized. What matters is if this will be just another boring spec upgrade, or will something new and exciting come forward.
How do you figure? I've been waffling about upgrading my 4S, and I'll make a choice when I see what Apple has in store. That said, the only complaint I have with the device is that some applications that leverage the larger screen size now aren't *as* usable as they could be. Still quite usable though. Until the battery stops taking a charge, there's little reason to upgrade; the thing does everything, and does it well. If I do upgrade, it'll replace my wife's BlackBerry Curve. Probably for at least a couple more years.
the ipad 5 would get the A7x in october
the iphone is apples flagship device so it comes first
Uhh, no, not at all.
The iOS kernel is based on the OSX kernel which has already gone through two 32->64 bit transitions, on both big and little-endian hardware. I would imagine there's very little of it that is not wrapped in macros which will automatically compile appropriately. There'll be the usual necessity for some manual intervention for VM and atomics, nothing that hasn't been done before.
The ARM A8 manuals describe how to run existing 32-bit binaries, and it doesn't involve emulation, any more than running 32-bit x86 binaries requires emulation. Finally, Apple is, you know, a SECRETIVE company. If they want to keep this secret until launch, they will not have this into the SDK. After all, it's not necessary right now.
The advantages of going to 64 bit for Apple are
- boasting. (Not technically important, but a way to impress the crowd who are excited by Moto's X8 architecture and Samsung's Octacore)
- allows them to move to >4GB of RAM EASILY and without a period of frustration. Apple has handled pretty much all these transitions well --- because they planned EARLY. MS and the Intel world have NOT handled them well (32->64 bit, various jumps in HD sizes, 4K disk sectors) because they DON'T plan years in advance.
- the sooner they go 64-bit the sooner they can use the A8 architecture. This is the real win, more so than a larger address space. The A8 architecture was ARM's chance to rethink everything they've done over the past ten years or so, improve what works well, and drop what was a bad idea.