Totally get this from a business perspective but speaking for myself, I was actually far more interested in seeing the new chip in the iPad. I have no doubt that the iPhone 12 will be another confident black rectangle. Having usb c and the stronger processor work its way down the price points was far more exciting of an announcement for me. I don’t really ever find myself yearning for a stronger CPU on my iPhone XR but if Apple is serious about replacing PCs with iPads, it seems like having strong chips in their cheaper models is a great position to fight from.
Likewise. I'm still rocking a 2016 13" macbook pro with a dual core CPU and integrated graphics. I'm looking at picking up this year's iphone, and it might be faster in every way than my laptop. (But with less RAM and storage)
If true, I'm not really sure what to do about that. I doubt I can offload rust compilation to an iphone.
Yeah, I have the 2015. You can actual compile Rust on an iphone (but I think you'd have to sync the code thei manually), but I think for Rust you really want one the AMD processsors with 8+ cores. I guess the new macbooks might be able to compete. We'll have to see.
>What applications benefit from this? The only use-case I've seen for AI in Apple's mobile devices is the face unlock.
Oh no. You could have double the CPU core and it wouldn't make 95% of common use case any faster. But NPU ( Neural Processing Engine ) and GPGPU have lots of potentials. Siri being the first one, When you take photos its uses NPU to adjust the image. Your Speech recognition for lots of different languages. Handwriting, Your Photo Gallery Face Recognition, on iPad Handwriting / Image to Text Recognition. And lots of thing likely to do with AR. And all of them has great impact to user experience previously not available.
Not to mention both GPU and NPU scales very well with transistor budget.
All inference on-device you mean? I presume that the vast majority of their models are trained in the cloud. Apart from some fine-tuning for personalization maybe.
While they did say machine learning, all the examples are of inference, and not training. So you don't have to ask if they meant inference when it is very clear what they mean.
AI creates possibilities of creating meaningful signal from extremely noisy environment, enabling lots of miniaturisation in sensors, which again enables tons of cross domain applications. e.g. low cost spectrometers for blood testing.
Speech recognition, image recognition, photo/video enhancement, face recognition, predictive text, etc. There aren't a huge number of applications, but there are enough to justify it, especially because it makes a big difference where it can be used.
It is worth pointing out A13 had a die size of 100mm2, much larger than usual, especially on a Smartphone Leading Edge Node.
So the A14 is basically am A13 5nm Die Shrink with better GPU and Double the NPU. What I am interested in is the Die Size. Which I am expecting to be sub 80mm2.
I'm not trying to be some sort of Luddite, but at this stage what more can a phone really do. If it's 10% faster than chips that are already doing a fine job right now it just feels flat.
I really love new tech and especially microprocessors but have trouble seeing how a phone breaks out of it's limiting form factor in any meaningful way right now that would make use of yet more processing power.
People have been saying that for 10 years. It makes calls, it’s got location, internet, a camera. Done right?
Since then we’ve got on-device powered AI voice assistants, Face ID, neural engine powered photo enhancement, health monitoring, custom silicon enhanced environment mapping AR. Goodness knows what else is coming.
The advent of custom tuned accelerators such as neural engines in silicon is opening up a huge new field of applications and capabilities. That’s especially true when they are linked to these new arrays of environment and bio-sensors. The X series iPhones are doing a list of things simply impossible on previous devices.
You may not care about some or all of these things, that’s fine they’re not all for everyone, but each one of them is important to some people.
Agreed - it’s the innovation in application of the technology rather than pure speed that has made mobiles far more interesting than desktops and laptops. Photography is the big ticket with most buyers now - the “how can I make every photo of my baby look like it is taken by a professional to my friends and family when it’s me just tapping once?”
I honestly think Apple could kill off all but professional photography and film with their own versions of photography kit such as Moment lenses, DJI Gimbals and associated AI correction tech to get rid of the shortcomings. Add in more powerful noise reduction in low light that their AI will soon enough bring and your phone camera is a better photographer than you will be without a great camera, training and practice.
The last real innovation in Intel chips in laptops and desktops I can remember that had a significant impact was the accelerated encryption and that’s years ago.
I’m honestly excited by Apple’s new chips going into laptops - Will we get webcam calls that have their image quality improved and background noise filtered by AI for example?
I know that all vendor spending much resources for improving cameras and works well, but I don't care about cameras on smartphone, so I feel smartphones are not evolving well nowadays. I wish they found new frontier other than cameras.
But how much does any of that really matter? I really believe the future is in things like AI voice assistance, but IMHO stuff like that isn't really good enough yet to make a big difference.
I think what Guthur may reflect upon is that we got tons of new phone models pushed into the market every single year that do very little new compared to previous models.
So what if we get 50% more performance. Sounds impressive, but then you realize it is only used to create smiley animations, and other frivolous teenager nonsense.
I don't think I have really seen a point in phone upgrades since iPhone 4 or so. That is about 9 years ago.
Same deal with iPads, not much new anyone needs. There are a couple of things though which I do think has mattered. Retina display and the Apple pencil.
But honestly the priorities are quite different from what I would want. I love my Apple products but I really wished they where a bit more modular or serviceable. By that I mean that e.g. replacing the battery should be much easier. Memory should have been easier to expand or replace.
Each year the companies listen to user feedback and iteratively improve the hardware and software.
I've never used an iPhone, but Galaxy phones used to have a really stupid and annoying button for their intrusive personal assistant. Now they don't have any such intrusive hardware or software designs.
The newest versions of Android, which are only supported on newer devices, allow much more fine-grained security permissions. In particular, I like that they allow blocking background access to a capability if the app isn't currently in use.
The new AR camera apps are really fun.
WiFi 6 and 5G provide noticeable improvement in internet speed.
The higher definition screens without bezels provide better video experiences for the same device dimensions.
Allowing fingerprint reading through the screen makes for a better experience.
Wireless charging, particularly in the car with a magnetic mount, is very slick and you never will want to go back.
If you need a cable, USB-C is clearly better than micro-USB.
All of this stuff only became common in the past couple years. If your phone is 3 years old, you probably don't have any of it. You will be fumbling with cables on a much less capable device.
I can't imagine what other sorts of polish they will figure out, but betting against it seems like a historical loser.
I still own an iPhone4S that I tried to use when there was an issue with the long-in-the-tooth (5 years old, 4 years newer than my 4S) but still perfectly supported 6S until it got fixed. It still runs iOS 6 which it came with (yes, that's a security risk I was willing to take for a couple of days).
It's not just the apps that are missing, and the low resolution; I preferred its theme, and the smaller screen but it's just not fun to use anymore now that I've gotten used to the snappy everything, the dictation while driving, the better camera, etc. When I got the 6S, it was for app support - I didn't feel anything was missing - but I won't switch back to 4S even if it did get a security update. I'll likely get a new iphone when iOS 15 comes out and the 6S isn't supported; There's nothing I know I need it for, but I suspect I won't be satisfied with the 6S after I do have it.
Re:ipads, though - my childern's 2013 ipad mini is essentially as good as their new one for all their uses, it's only missing software and security updated that prompted getting a new one.
6S felt like a quantum leap from every other phone before it and I’m still using one too. I can’t nail it down but there was something perfect about the size, speed, UI where using a phone stopped feeling like messing with a gadget and became pure information retrieval or communication. Trying every other iphone afterwards I have felt no need to upgrade.
I found my old iPhone 3GS is a drawer in 2016. Not only did it still work fine, but it even connected to the App Store and you could still download games I’d had back in the day that werent on the phone, even though those games couldn’t run on any current iPhones. That’s a full seven years after I bought it.
I have an Ipad mini (gsm model) and most of the apps are not working/updating anymore. E.g whatsup, facebook etc. I believe the "mail" app is not working anymore either. I guess the games(most of them) do not depend on the ios sdk so that could be a reason why they still work.
There are two types of compatibility. With the OS and with external network services that may have changed API. You won’t be able to run apps built for later os versions than you have on the device, but these just won’t show up in the App Store for you. Only older versions of the apps compatible with your os.
Things like Facebook and WhatsApp apps depend on the developer to keep the app compatible with their network API. I would expect mail to still work, depending on which service you use. Any custom hooks into GMail for example might fail if Google changes the back end interface.
The apps themselves should still ‘work’ as in run on the device just fine as they were built for that OS, it’s compatibility with external services that may have changed their APIs that is the main issue.
In practice Apple forces the developers to use only new APIs(available on new OS) while the device is not supporting new OS updates.
Submissions with the old APIs are no longer accepted so tell me, isn't that planned obosolence?
Basically I as a developer and owner of the device cannot update my own app, on my old device anymore because the OS does not support the new APIs and Apple does not allow me to submit updates with the old APIs or to upgrade the OS.
Your only choice is to not use Apple APIs that are changed often. In practice you can do that only on certain games or niche apps.
Your contradicting yourself. How can they be supporting their devices for longer (much longer) than their competitors because their premium customers expect it, but also guilty of dropping support for the same products early to make more money? You’re not making any sense.
So you think it's all right to have your computer become a useless brick after 5 years because the other vendors makes it a brick after just 2 years?
It's also worth to note that at least on Android you can still side-load/install apps outside the play store while on Apple your device becomes totally useless/museum brick.
The number of competitors has little to do with planned obsolescence but
" Planned obsolescence tends to work best when a producer has at least an oligopoly "
So our quibble is over “useless brick.” How does Apple dropping support for a device make it a “useless brick”? As far as I’m aware, you can still use an original iPhone from 2007 today. It won’t be fast, but it will work more or less.
Planned obsolescence is not related to advancements in technologies. Would you classify the Commodore 64 as having been designed with planned obsolescence in mind considering it ran on a 6502 when the competition at the time was transitioning to 8086? Of course not. So why is Apple deciding to put more advanced technologies into their devices planned obsolescence?
If you’re referencing the App Store, though, I can see where you’re coming from. But Apple releasing new technologies and not putting them on devices that can’t handle them (due to computing power) and developers using them (which prevents them from running on said old devices) is not planned obsolescence either. The developer chose to make it incompatible with the older devices, not Apple.
I'm not trying to use new tech on old devices. I just want to update the apps using "old tech"/sdk (which is the only one that works) but Apple store doesn't accept old tech/code regardless the device you try to target.
To answer your question:
today I can still deploy code on Commodore 64 but not on my old iPhone. How is that possible?
The same can be said for old consoles, or old non-smartphones that were also computers, had browsers, even in some cases have CPUs more powerful than contemporary iPhones, but didn't have any way to install apps at all.
There's no law requiring vendors to provide features to install apps on devices with CPUs at all.
You probably own a dozen devices containing CPUs, running software, for which there is no practical way for you to install anything. Apple provides a certain set of features in their devices, if you don't think that's a good deal, there are other vendors you can go to.
That is correct, the assistant needs to be able to retrieve data and assets it needs from back end services to be able to perform efficiently. So what?
Even from your perspective—which isn’t Luddism, it seems to be an assumption that a small handheld device with passive cooling is inherently underpowered and that larger form factors will always have a performance advantage—the thing that phones are able to do right now, at least in Apple properties, is improve the performance/power ratio so significantly that one of the largest technology brands is switching their entire large form factor lineup to the chips designed for their small handhelds.
In other words... the thing phones can continue to do for general processing power is dictate efficiency by necessity, and make that efficiency generalizable if there’s an interest in that investment.
Consider that these more powerful phones may become servers for new wearable form factors that aren't possible to make with their own chips. Smart glasses come to mind, the phone can do all the image processing over a wireless signal.
Another interesting possible future is the desktop replacement, dock your phone into a hub that exports video and USB signals to connect a keyboard mouse and screen and do away completely with laptops for user's whose computing needs are modest. Samsung does this already with DEX but it's too limited for me to view as more than a cool prototype but with Apple's app ecosystem this could be quite useful.
All type sensor, spectrometers, blood testers, AI related medical apps, 3D scanning, etc etc will need high performance platform much before actual application available.
The number of cameras on phones will continue to grow, and soon they will fuse data from multiple sensors at a time to produce one image or video (and not just in portrait mode), like this camera from a few years ago:
> and soon they will fuse data from multiple sensors at a time to produce one image
The iPhone already does this, outside portrait mode. It’s called Deep Fusion and triggers on well lit photos shot using either the 1x or 2x lens (aka not on the 0.5x, super wide, lens).
If it doesn't go towards making it beefier, it will go towards making it smaller. Ambient computing (VR glasses or optical/cochlear implants) seems to be one logical terminal point. Just like the "PDA" has now become an outdated form factor, so too will the phone.
For phones, I probably agree at this point. I'm very happy with my 11 pro max. It's fast, takes great pictures for its size, and I'm never waiting for it to do something. At this point more efficient chips could be used to improve battery life, but that does not always seem to be a goal. I'm sure Apple (and others) has some research somewhere that shows most people need X hours and that is their target.
My 2018 iPad pro is sort of the same, but as I find myself using it for more and more things, I can see places where more speed would help. It's my goto device now for using LR and sorting and editing pics from my DSLR. It's already amazingly fast, but in cases like this more speed is still generally better.
These advancements wouldn't matter to someone who just bought a device last year but like Bill Gates once said, it pays off in the long run. So for someone who is upgrading after a few years, they'll notice it, and they'll continue to notice it for next few years before they make another leap.
I read comments on the benefit of real world changers like new health sensors (or sensors of any type now already on phones) but the way I see, this additional processing power is being used mostly at:
- "AI" photography. Trying to make pictures looking like it was taken by a non-mediocre camera
> two high performance A13 cores are 20% faster with 30% lower power consumption than the Apple A12's, and the four high efficiency cores are 20% faster with 40% lower power consumption than the A12's. A13's eight-core Neural Engine dedicated neural network hardware is 20% faster and consumes 15% lower power than the A12's
> A14 up to 40% faster than A12, with 30% faster graphics performance than A12, and machine-learning performance is up to 10 times faster.
TSMC says that moving from their 7nm to 5nm process node offers either a 15% speed improvement or about a 30% reduction in power consumption.
The thing that I found interesting in the A14 is that this is the first time I can remember a chip designer using a die shrink mainly to cut power usage instead of increase performance.
However after Anandtech's recent deep dive into Tiger Lake performance testing, you can see where cutting power use in the A14 cores might be a winning move.
>Here we present the 15W vs 28W configuration figures for the single-threaded workloads, which do see a jump in performance by going to the higher TDP configuration, meaning [Tiger Lake] is thermally constrained at 15W even in ST workloads.
Comparing it against Apple’s A13, things aren’t looking so rosy as the Intel CPU barely outmatches it even though it uses several times more power, which doesn’t bode well for Intel once Apple releases its “Apple Silicon” Macbooks.
> The thing that I found interesting in the A14 is that this is the first time I can remember a chip designer using a die shrink mainly to cut power usage instead of increase performance.
This has been a thing with RF Chip design for the 30 years I've been paying attention. Transceiver I used 20 years ago consumed 220 mW in receive mode. One I used 10 years ago, 50 mW. Current one uses 10mW.
I can't come up with a single other example where cutting power was prioritized.
Things have tended to go the other way with die shrinks that favor performance and often with ridiculous amounts of overclocking piled on top of that.
>Brian and I have long been hinting at the sort of ridiculous frequency/voltage combinations mobile SoC vendors have been shipping at for nothing more than marketing purposes. I remember ARM telling me the ideal target for a Cortex A15 core in a smartphone was 1.2GHz. Samsung’s Exynos 5410 stuck four Cortex A15s in a phone with a max clock of 1.6GHz. The 5420 increases that to 1.7GHz. The problem with frequency scaling alone is that it typically comes at the price of higher voltage. There’s a quadratic relationship between voltage and power consumption, so it’s quite possibly one of the worst ways to get more performance.
It's a bit strange in the same year they're going to present the first laptops sporting the new ARM chip. I mean the performance increase can be declared by the clock increase the smaller process allows alone. So either they put all of their engineering capacity towards a blistering fast desktop / laptop CPU or they really have a problem in their SoC engineering department.
Also they reserved all of TSMC's 5nm capacity for the A14. Like that was a last ditch effort to make it _look_ faster after all?
The upcoming Snapdragon and Exynos chips are on Samsung's 5nm process, so they should be relatively comparable. I wouldn't worry on that front.
> So either they put all of their engineering capacity towards a blistering fast desktop / laptop CPU
TSMC promises up to a 30% improvement in power consumption or up to 15% more performance over the current N7 node in the 5nm process. This relatively matches with Apple's claim this gen.
So Apple might have put the engineering effort on desktops & laptops this time.
Transitioning to a new node can be tricky, so in the past they have been very conservative in their changes to the design when doing so. Better to do the transition with as few new potential surprises as possible.
This is also true for Qualcomm so I’d expect to see similar, mainly process node driven changes in their new chips as well.
