Apple's Federighi answers developer's questions on Siri, Apple Silicon, and more

Posted:
in General Discussion edited November 2020
Apple SVP of software engineering Craig Federighi has appeared on Marques Brownlee's podcast, answering questions that developers have raised since the WWDC keynote.

Apple's Craig Federighi
Apple's Craig Federighi


After his and Apple marketing vice president Greg Joswiak talked on video, Craig Federighi has now also appeared on Marques Brownlee's Waveform podcast. He was questioned about Apple's thinking over the technologies announced at WWDC, and also pressed on criticisms that developers have had since then.

Siri's new interface

One of those concerned the new visual form of Siri, which no longer covers the entire screen -- but which does prevent you from interacting with what else is on that display at the time. Federighi says that's intentional.

"We tried it both ways," he said. "Internally, we had prototypes running, where you would scroll behind the Siri answers as you brought them up. And initially we thought that was going to be really great, so we implemented it that way and then what we found was that for most of the interactions we were having with Siri, that then created this kind of extra heaviness to it."

"Because if you just say, wanted to check the weather, or look up a piece of information, and then you were ready to move on, your scroll gesture wasn't gonna move on," he continued. "The Siri result was going to stick around there so you always have this additional step of dismissing the result."

"We'll continue to listen to what people say through beta period because, like I say, we've actually had it working both ways," he said. "But our feeling was that we wanted to achieve a great lightness, not just visually but in terms of you could dip into Siri get an answer and move on quickly without any kind of overhead."

Default apps

Federighi also addressed the question of why Apple was now allowing the default Mail and browser apps to be changed by the user, but nothing else, no other apps. He said that the choice for now was deliberate, but that didn't mean it won't change.

"We know how platforms can sort of descend into chaos," he said, "and particularly when apps that you know might honestly not even be browsers... decide they're going to be a browser. And then [they] try to redirect you to a different tract ad experience when you click on something... or who knows what."

"And so we are very careful about making sure that that experience that we don't have people misled," he continued, "or have their device gets configured into a place where boy they just don't even like like using their phone anymore because it's gotten so locked up. So we proceed cautiously on these things for sure."

The new look of Siri in iOS 14
The new look of Siri in iOS 14

Apple Silicon and apps

Apple has been criticized before for saying that developers can just use a tickbox and have their apps convert to a new platform or some new system. So Brownlee pressed Federighi on just how true it can be that developers will find converting to Apple Silicon easy.

"We feel really great about how the transition is already starting and how it will go," started Federighi. "Most applications out there have really modernized over the years to use our latest developer tools, to be 64 bit clean, [and] to use our modern set of frameworks and it turns out that that means moving to Apple Silicon in some cases is literally recompiling."

"We've seen many examples of that," he continued. "More sophisticated apps like Microsoft Office and Adobe Suite, some of these apps have decades of engineering, you know, hundreds and hundreds of people working for decades creating code and sometimes having dependencies on different libraries and so forth and those are bigger projects."

"But I think the amazing thing that I hope everyone can take away from the presentation was that those apps, [Office and Adobe], have been completely brought over," he said.

"And it wasn't like [Microsoft or Adobe] had to mobilize the entire engineering team," he continued. "This was incredibly secret, and so we said, could you give us a person or two and let's get those apps up and running. "

"We [did] it in a relatively short period of time and those are some of the hardest apps you could imagine bringing over," he said.

The full 90-minute podcast can be heard here, and includes much more about Federighi's take on WWDC, and especially the exceptional video keynote.

Keep up with all the Apple news with your iPhone, iPad, or Mac. Say, "Hey, Siri, play AppleInsider Daily," -- or bookmark this link -- and you'll get a fast update direct from the AppleInsider team.
watto_cobra

Comments

  • Reply 1 of 10
    Based from all that have seen since Monday I think this is what went down a couple years back in Apple Senior officer meeting. 

    “Ok guys, we have an important decision to make. Our laptops are burning. We can’t control the heat. Our design roadmap is no longer applicable with current CPU. The damn thing heats up too much. We are getting so much heat from our customers. Our engineers can’t keep up with finding thermal solutions to mitigate without doing the whole chassis revamp. It’s screwing up our product launch plans. We’ve complained to them [Intel] and are not getting any resolve. What can we do?”

    ans: “we can use our own chip!” Saids Johnny Srouji

    ” yeah. But is our chip good enough to replace Intel? This decision can not be a decision based on anger. If we are gonna move, we have to know that what ever we have has to be better. Otherwise what’s the point? “

    Ans: yes. Our solution is good. My team and I have been working on monolithic prototypes of the chip. And it’s good. I mean really good. We can fit it inside the chassis with minimal thermal. It runs like a beast and the graphics is so great.

    ”but if we move to ARM, how about x86 GPU? Our customers are gonna be expecting the performance that AMD will bring. Can we even do that?”

    ans: yes. Our prototype chip is a scaled A13 chip. We know what Nvidia and AMD are doing with their upcoming offerings. What we can do is increase our compute units but keep it at low power consumption. The result is a low heat but powerful chip that matches a Dedicated GpU solution. Our team uses the dedicated GpU from AMd and Nvidia as benchmark for development criteria’s. To convince our customers to move to ARM with us is a major move. So the bare minimum of performance must be set higher than what will be available when and if the chip launches. 

    “So. Can we do it? And are we going to do it?”

    ans: [everyone] yes. Let’s do it. 

    - - -
    this is speculative. But it’s the sort of discussion that we do at my company. We often have supplier reviews. Check their quality and delivery performance. We discuss long term strategic plans like this as well. Weight out the pros and cons. Then make the decision and announce. We are not a big company like Apple, but I expect their senior discussion meeting to be quite similar. Their decisions have many ramifications. To their users, suppliers and developers. Their decision to move must have enough reasoning to justify the risks taken. Remember, the said chip, is guarded behind the mothership’s walls. Basing anything on just the Dev Kit A12z  as benchmark isn’t a good idea. Strategically speaking, they know they need to release an ARM chip to the public so they can write their code on. The A12z is already good and is publicly known. “We’ll use that for developers so they have a test platform. We can’t release the new chip just yet. It’s gotta be kept secret for the time being”. I think Apple knows regardless of their “no benchmark” policy of the dev machines, someone will still do it. Words gonna get out how well or bad the chip is gonna do on MacOS Big Sur. So giving devs / public a glimpse of “hey! A12z is a great chip right? It can do anything you ask of it.” 

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    edited June 2020 watto_cobra
  • Reply 2 of 10
    michelb76michelb76 Posts: 659member

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    Windows exists on ARM, but many Windows software does not support it. So like most scientific software that has moved away from the Mac, there will be more software unavailable for Mac users, even with VM capability available. This will mean a lot of students will now have to buy a Windows laptop. Games will not be playable under Windows on the Mac since they are not playable on Windows ARM either.

    It will be interesting to see if they will make a CPU that will be relevant for use in the MacBook Pro. And I'm very interested in developer tests with regards to build times.


  • Reply 3 of 10
    blastdoorblastdoor Posts: 3,431member
    I think Apple has a very good chance of taking the crown from x86 both in single thread and multi-thread. Three reasons:

    1. Apple is TSMC’s number on customer. They are a year ahead of AMD in accessing the latest fab process

    2. Apple has far more resources that AMD for design 

    3. apples control of the entire software stack gives them an advantage in optimization 
    jdb8167thtwatto_cobra
  • Reply 4 of 10
    darkvaderdarkvader Posts: 1,146member
    michelb76 said:

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    Windows exists on ARM, but many Windows software does not support it. So like most scientific software that has moved away from the Mac, there will be more software unavailable for Mac users, even with VM capability available. This will mean a lot of students will now have to buy a Windows laptop. Games will not be playable under Windows on the Mac since they are not playable on Windows ARM either.

    It will be interesting to see if they will make a CPU that will be relevant for use in the MacBook Pro. And I'm very interested in developer tests with regards to build times.



    They don't stand a chance of taking any crown.

    ARM/Awhatever is "good enough" in mobile devices.  It's not good enough for a real computer.

    Apple has made a MASSIVE mistake, and they're about to lose a lot of customers.  Virtually everyone who moved to the Mac after the Intel switch because they could still run a few critical Windoze apps WILL be leaving.  Thousands of IT departments that allowed the Mac will be banning it again.  Developers will abandon the platform.  The already mediocre Mac game scene will evaporate, virtually the only games on the platform will be iOS games.

    Apple just killed the Mac.  The death throes will last years, but they're already beginning.
    xixo
  • Reply 5 of 10
    darkvaderdarkvader Posts: 1,146member

    “Ok guys, we have an important decision to make. Our laptops are burning. We can’t control the heat. Our design roadmap is no longer applicable with current CPU. The damn thing heats up too much. We are getting so much heat from our customers. Our engineers can’t keep up with finding thermal solutions to mitigate without doing the whole chassis revamp. It’s screwing up our product launch plans. We’ve complained to them [Intel] and are not getting any resolve. What can we do?”

    "We can make them thicker.  It's what our customers want anyway, almost nobody really cares about having a laptop that's a quarter inch thick.  We can keep making thin and underpowered 'MacBook Executive' models for the few idiots who do, and sell them for stupid money.  For everybody who actually needs a computer to use, we can make the MacBook Pro an actual professional laptop again.  Oh, and why are we still exclusive with Intel?  AMD has some nice x64 chips now.  Let's use those too.  Our customers will love us!"

    I mean, that's how it obviously should have gone.  Instead they took the stupidest route possible.
    xixocommand_f
  • Reply 6 of 10
    canukstormcanukstorm Posts: 2,729member
    darkvader said:
    michelb76 said:

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    Windows exists on ARM, but many Windows software does not support it. So like most scientific software that has moved away from the Mac, there will be more software unavailable for Mac users, even with VM capability available. This will mean a lot of students will now have to buy a Windows laptop. Games will not be playable under Windows on the Mac since they are not playable on Windows ARM either.

    It will be interesting to see if they will make a CPU that will be relevant for use in the MacBook Pro. And I'm very interested in developer tests with regards to build times.



    They don't stand a chance of taking any crown.

    ARM/Awhatever is "good enough" in mobile devices.  It's not good enough for a real computer.

    Apple has made a MASSIVE mistake, and they're about to lose a lot of customers.  Virtually everyone who moved to the Mac after the Intel switch because they could still run a few critical Windoze apps WILL be leaving.  Thousands of IT departments that allowed the Mac will be banning it again.  Developers will abandon the platform.  The already mediocre Mac game scene will evaporate, virtually the only games on the platform will be iOS games.

    Apple just killed the Mac.  The death throes will last years, but they're already beginning.
    Hyperbole much?
    jdb8167fastasleepwatto_cobra
  • Reply 7 of 10
    xixoxixo Posts: 449member
    my question: when will I be able to install Linux on my iMac Pro's internal SSD and boot from it?

    after all, if Intel MacOS is going away, I need to be able to consider other options.
  • Reply 8 of 10
    michelb76 said:

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    I've seen people saying a lot of things like that about unreleased/just announced Apple products... quite more often than not, they shove their own feet, knee deep, into their mouths.

    It would be no surprise, at all, to myself if Apple were to fundamentally redefine capabilities and performance per watt figures for SoC.
    jony0fastasleepwatto_cobra
  • Reply 9 of 10
    darkvader said:
    michelb76 said:

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    Windows exists on ARM, but many Windows software does not support it. So like most scientific software that has moved away from the Mac, there will be more software unavailable for Mac users, even with VM capability available. This will mean a lot of students will now have to buy a Windows laptop. Games will not be playable under Windows on the Mac since they are not playable on Windows ARM either.

    It will be interesting to see if they will make a CPU that will be relevant for use in the MacBook Pro. And I'm very interested in developer tests with regards to build times.



    They don't stand a chance of taking any crown.

    ARM/Awhatever is "good enough" in mobile devices.  It's not good enough for a real computer.

    Apple has made a MASSIVE mistake, and they're about to lose a lot of customers.  Virtually everyone who moved to the Mac after the Intel switch because they could still run a few critical Windoze apps WILL be leaving.  Thousands of IT departments that allowed the Mac will be banning it again.  Developers will abandon the platform.  The already mediocre Mac game scene will evaporate, virtually the only games on the platform will be iOS games.

    Apple just killed the Mac.  The death throes will last years, but they're already beginning.
    It's funny how people have so many definitions of "real computer". To me (I'm a professor/researcher at academia) it's a HPC platform (dozens to hundreds of Tflops/s), which has never been a Mac (Intel based or otherwise), although there are more than a few developments with Swift+Metal on that regard, particularly with the new Mac Pro.

    To some, who would like to appear knowledgeable about imagined future IT policies, a real computer is all about games... don't get me wrong, I enjoy games as much as the next guy, but that has never been a tie breaker on which piece of hardware I've ever bought. My Mac, iPad and iPhone were bought with work considerations in mind.

    If a developer house, particularly for games, choses to eschew macOS, iPadOS, iOS, and tvOS (I'll leave watchOS out of this) platforms... that's their choice. I'll never be so bold as to declare their death in a few years time with such an absolute certainty. And I certainly would never say that of such a company as Apple! The only major computer hardware company to have ever succeed in a platform migration (twice!!!).

    As form x86(-64) itself, I say good riddance. It's long overdue... Intel itself tried to move the market away from it (remember Itanium???), being quite aware of x86 architectural bottlenecks. Then leading PC manufacturer, HP, among other, betted heavily on it, only to their collective sorrow. Market inertia gave then no room to breathe... luckily (even though luck has nothing to do with it), this does not seems to have ever bothered Apple.
    jony0command_fwatto_cobra
  • Reply 10 of 10
    fastasleepfastasleep Posts: 6,447member
    darkvader said:
    michelb76 said:

    Decoy tactics. Cause Intel / AMD / Nvidia / Qualcomm are laser focused on what Apple will do next. Why would Apple handedly let the cat out of the bag before prime time? If I’m gonna launch my own chips, why would I let the competition know? I’m gonna trick them into thinking that our chips are weak. Then, launch a surprise. Shake up the market. Take the crown for my self. 

    What Craig said, pretty much answered most of the worries. It supports what I just described. You guys should watch it. 
    I doubt they will take a crown, maybe in the watt/perf portion of the market. Both Intel and AMD suffer from the same architectural problems, although AMD is actually trying something. For the foreseeable future (several years) Apple will not have a chip that rivals the higher-end consumer Intel/AMD chips. So it obviously first targets MacBook and iMac users who don't care about performance. A great testcase though. Keeping the beefy Intel CPU's for the Mac Pro, and iMac 'pro', although that will be even worse value for money in a year from now. 

    Windows exists on ARM, but many Windows software does not support it. So like most scientific software that has moved away from the Mac, there will be more software unavailable for Mac users, even with VM capability available. This will mean a lot of students will now have to buy a Windows laptop. Games will not be playable under Windows on the Mac since they are not playable on Windows ARM either.

    It will be interesting to see if they will make a CPU that will be relevant for use in the MacBook Pro. And I'm very interested in developer tests with regards to build times.



    They don't stand a chance of taking any crown.

    ARM/Awhatever is "good enough" in mobile devices.  It's not good enough for a real computer.

    Apple has made a MASSIVE mistake, and they're about to lose a lot of customers.  Virtually everyone who moved to the Mac after the Intel switch because they could still run a few critical Windoze apps WILL be leaving.  Thousands of IT departments that allowed the Mac will be banning it again.  Developers will abandon the platform.  The already mediocre Mac game scene will evaporate, virtually the only games on the platform will be iOS games.

    Apple just killed the Mac.  The death throes will last years, but they're already beginning.
    Yeah, or virtually none of that is true.
    watto_cobra
Sign In or Register to comment.