Intel Expecting Apple to Transition to Custom ARM-Based Chips Starting in 2020

[0837990]

iOS niche platform? What the hell are you smoking? Market share in and of itself is a useless metric.
[doublepost=1550835199][/doublepost]
I have an iPad Pro too .

IOS is becoming a niche platform compared to other, much larger platforms. Many IOS developers have changed focus, and are now focusing more on other bigger platforms. Why? Simply because the much bigger platforms is where the money and income is. It used to be IOS/AppStore that generated most of the app revenues, but this has changed in the last years, and revenue-wise, IOS is "dying" compared to the same apps on Android. Even dedicated IOS developers that put effort into porting their apps to Android are seeing significant larger revenues on Android these days. And the result of that should be fairly simple to see... The developer(s) will focus more on the platform that generates the most revenue. Of course!

And not only that, Google's Play Store is a LOT easier to work with than Apple's AppStore for developers. The AppStore, and all it's guidelines, rules, regulations, and policing has become a royal pain in the but* for developers to deal with... I know some developers who has worked for weeks and months on just getting an app out on the app store, simply because Apple's process was a royal PITA and took forever to go through. That also makes developers not wanting to work with the platform.

If anything, this change might result in it being a pain to be a macOS developer. 95% of my macOS apps are not installed via the AppStore, and I'd prefer it stays that way for the foreseeable future.

 

[timber]

Besides that holy grail of performance benchmarks, Geekbench, which other tests compare x86 offerings to Apple's A series?

It will be a pain but there is no reason why A designs won't best intel. Intel doesn't even have its process advantage that used to carry it in the past.

 

[dilbert99]

I hope Apple moves to ARM soon. I believe Intel is yesterday's news. The ARM processors are getting very fast and the graphics capabilities are better than what we see in Intel-based thin and light laptops with Intel Iris Pro or Intel HD graphics. My MacBook doesn't have the muscle my iPad Pro has both in CPU and GPU performance. It's time to ditch Intel and move to ARM. Bring it on!!

Armchair reporting is all the craze nowadays...

 

[Tiger8]

This will be as good as Apple maps

 

[dilbert99]

Whats so impressive about an i7? I bought a brand new iMac last summer with one and it feels more sluggish than my 2013 and has more issues. In the past, a 3+ year gap between computers was a wayyy bigger upgrade

And you still kept it rather than take it back?
[doublepost=1550838134][/doublepost]

Duh, the iPad Pro with its A12x outperforms Intel’s mobile processors.

Is that while running a single app or 10 apps at the same time?

 

[timber]

This will be as good as Apple maps

Apple Maps sucks but not everything Apple does sucks (well most things really rock).

We can hope it would be like the A chips. Oh wait, these are the A chips.

 

[baryon]

So does this mean that starting in 2020, we'll have to live through the same kind of annoying mess that was the transition from PPC to Intel, such as having "Universal" apps, Apps that only run on ARM, apps that only run on Intel, and apps that refuse to be updated (Adobe) for a long time, apps that you prefer to use the old version of but only the new version will run natively on your CPU, apps that run through some sort of abstraction layer like Rosetta and are much slower than native apps, apps that won't get ported over ever and will just die, apps that get ported over but lose a bunch of features, etc...?

 

[TheRealNick]

This will surely lead to a locked down eco system on the Mac.

And since the majority of the software I use on Mac is not in the App Store, it will make the decision between purchasing a Mac and Windows (ugh) machine more difficult.

Either I use the software I need in a terrible operating system, or I use poor alternatives (or in some cases just go without) in a stable, nice to use operating system.

 

[Michael Scrip]

This will surely lead to a locked down eco system on the Mac.

And since the majority of the software I use on Mac is not in the App Store, it will make the decision between purchasing a Mac and Windows (ugh) machine more difficult.

Either I use the software I need in a terrible operating system, or I use poor alternatives (or in some cases just go without) in a stable, nice to use operating system.

Come to the dark side.

It's not all bad... you might get a good laptop keyboard!

 

[crsh1976]

Hopefully they will not move the entire line to ARM, dropping x86 compatibility ane current pro software is a sure way to make us drop Macs altogether across the whole company.

But hey, it's not like Apple actually gives a crap about the pro market.

 

[Heineken]

So does this mean that starting in 2020, we'll have to live through the same kind of annoying mess that was the transition from PPC to Intel, such as having "Universal" apps, Apps that only run on ARM, apps that only run on Intel, and apps that refuse to be updated (Adobe) for a long time, apps that you prefer to use the old version of but only the new version will run natively on your CPU, apps that run through some sort of abstraction layer like Rosetta and are much slower than native apps, apps that won't get ported over ever and will just die, apps that get ported over but lose a bunch of features, etc...?

I remember it being ok. No big discomfort. I had an iMac G4 and replaced it with MacBook core 2 duo. No big deal.

 

[hagjohn]

I don't think Apple would consider switching unless they knew their chips could compete. We lived through it once, I'm sure we will all survive. --- Me, not being the negative guy...

 

[Heineken]

I don't think Apple would consider switching unless they knew their chips could compete. We lived through it once, I'm sure we will all survive. --- Me, not being the negative guy...

Two A12X are faster than an i7 for a fraction of a cost. That’s guaranteed. Nobody knows what designs they have in the labs.

 

[justperry]

Mac Pro, iMac Pro.

Current Mac Pro?

Oops, I forgot about it/didn't realise/need coffee.

 

[LordVic]

I would be very interested to see if this actually materialises. Most users here don't truly understand what the deal is and comparison points are. They just think benchmark results are easily comparable.

They also think that if you just threw more wattage through an A series CPU it would automatically be as fast if not faster than the x86 products, not taking into account that it doesn't scale that way and Apple, or any arm manufacturer are going to run into the same laws of physics everyone else will.

I think what you'll find if they go ARM is that there will be some benefits, some stuff will feel snappier, but some activities just won't run well. There are benefits to both platforms, and so far, high power performance has not been an ARM one. Apple's ARM CPU's are absolutely the best in the ARM business. We often compare them to Intel's similar watt Y or U series CPU's and think that if they're as fast as a 7w i7 they must be as fast as any i7. This is a logical fallacy as they are not created equally.

Either way. if Apple goes ARM only, it will probably be the last time I ever look at an Apple computer (although, they've already dropped significantly on my list due to the current offerings). I have nothing personally against ARM, it just will not work into my workflow which does require windows/linux with legacy support.
[doublepost=1550842353][/doublepost]

Two A12X are faster than an i7 for a fraction of a cost. That’s guaranteed. Nobody knows what designs they have in the labs.

Really? you think Apple's A12 is faster than an i7? or, are you just picking and choosing WHICH i7 you want to compare to?

https://cpu.userbenchmark.com/Compare/Intel-Core-i7-9700K-vs-AMD-A12-9800-APU-2016-MBR/4030vsm283761

This is the competition Apple has to match/beat for the transition to be truly effective.

 

[frou]

Fast-forward 10 years and thinking about the bombastic x86-or-bust loyalists will be a giggle.

 

[cmaier]

Fast-forward 10 years and thinking about the x86-or-bust loyalists will be a giggle.

They’ll be clinging to their 2020 MBPs, exchanging tips on how to replace malfunctioning keyboards and links to home brew web browsers that work with Wi-Fi 12.

 

[ahireasu]

Interesting, how is it you know so much about what Apple has in the lab? Their publicly visible work on the ARM platform has been optimized for making phones perform extremely well on low power. Those processors, running in an extremely compact environment, outperform a lot of laptops. When they go to ARM for the desktop, it's unlikely they'll just take their phone CPUs and use them unchanged in laptop/desktop machines. They'll do it after working up chips with much higher performance, to work within the power/thermal environment of a laptop or desktop, rather than a phone.

So are you saying that apple will make desktop Gpus that will rival THE GTX 2080 of the future?or that will create an ARM Cpu with 32 cores etc?
I know apple was always greedy and cheap in putting dedicated Gpus,was/is slow to update their hardware with the latest and greatest.
I don't need to be inside apple to know all I need to do is look the past 10 years or so.

 

[ocnitsa]

Lots of talk about Apple merging OSs to achieve something unified...has anyone thought about how laptop form factors may be eliminated or may change? I realize that we are talking about universal apps...but surely this is a move to a universal OS among Apple devices (at least phones, tablets, computers), no? I know that many don't want to imagine fundamental changes, but Apple seems unafraid to make such changes (like the shift to USB-C with no legacy port support). With a universal OS, if Apple decides that mouse support is a thing of the past (as they've stated in regards to iOS), there is no need for a laptop form factor...The same function would be achieved by iPads (maybe with more beefier CPUs) and smart keyboards (with Apple pencils). If this is the case, I can see Apple raising prices from price points of previous iPads...much like iPhones, they will argue these things do so much more now and fulfill multiple functions. It's possible I'm wrong here, but I think Apple is working towards a unified version of hardware, too, at least in terms of phone, tablet, and computer. I think this would lower the cost for manufacturing, because there would be less variation in products, but, as I said, I think Apple would find a justification to raise prices. Ive has said his end goal for the iPhone is a single slab of polished glass. If all apps run on iPhones, iPads, and laptops, then there is no reason to need laptops anymore...no reason not to imagine a single slab of polished glass.

(Just want to add that I am not on board with this idea...I prefer a Mac that includes bootcamp. Also, prefer a keyboard I don't have to worry about breaking down.)

 

[gnipgnop]

Really? you think Apple's A12 is faster than an i7? or, are you just picking and choosing WHICH i7 you want to compare to?

https://cpu.userbenchmark.com/Compare/Intel-Core-i7-9700K-vs-AMD-A12-9800-APU-2016-MBR/4030vsm283761

This is the competition Apple has to match/beat for the transition to be truly effective.

Do you think that Apple is unaware of what a real-world comparison of x86 and ARM would be? There's no doubt that Johny Srouji and his teams would understand the differences and challenges there. If Apple did make a transition like that, then there is going to be a LOT of research/development time behind it. Remains to be seen if that's what they're going to do.

 

[cmaier]

Do you think that Apple is unaware of what a real-world comparison of x86 and ARM would be? There's no doubt that Johny Srouji and his teams would understand the differences and challenges there. If Apple did make a transition like that, then there is going to be a LOT of research/development time behind it. Remains to be seen if that's what they're going to do.

They won’t even consider it a challenge. It’s just another task. “You want double the performance with x percent higher tdp? Ok. We will get started on that.”

It’s just as hard or harder to design for phones as for desktops. The constraints are different in each case, but the same tools and same bag of tricks are available.

Hand me a transistor spice deck, a power budget, a die area budget, and the DRC deck, and I’ll get started.

 

[Falhófnir]

Really? you think Apple's A12 is faster than an i7? or, are you just picking and choosing WHICH i7 you want to compare to?

https://cpu.userbenchmark.com/Compare/Intel-Core-i7-9700K-vs-AMD-A12-9800-APU-2016-MBR/4030vsm283761

This is the competition Apple has to match/beat for the transition to be truly effective.

I mean, considering it's a 7.5W passively cooled chip, I'd be more inclined to compare the A12X to intel's core M chips - which if we're being polite we could say it compares favourably to? Even if performance doesn't scale linearly with increased wattage and frequency, I think it's fair to say Apple wouldn't be too hard pressed to match the performance of the chips you've linked.

 

[LordVic]

Do you think that Apple is unaware of what a real-world comparison of x86 and ARM would be? There's no doubt that Johny Srouji and his teams would understand the differences and challenges there. If Apple did make a transition like that, then there is going to be a LOT of research/development time behind it. Remains to be seen if that's what they're going to do.

Absolutely, But they have an uphill battle to climb and it's not easy (even intel is currently running into physics barriers with performance).

the thing about the PPC to x86 transition that helped was that the power and efficiency was so much great for x86 that for several iterations Apple was capable of emulating PPC to allow for a smooth transition.

the x86 to ARM transition would be a lot more disruptive as the ARM CPU's are not showing that they have enough headroom for this in full. Current attempts are showing promising (Windows 10 showcased on ARM running 32bit apps in emulation have been moderately succesfull from a proof of concept). But we haven't yet seen ARM full blown emulate full x64 programs.

I think what is more likely is that Apple will migrate the MacBook to ARM as a very low power laptop with 2-3 day battery life, but will likely leave the MacBook Pro, Mac Mini and iMac on intel.

 

[rlhamil]

This will be a disaster. The virtual machines that I run are going to crawl under any kind of emulator that Apple produces, and I'm not all that sure they'll even throw us that bone. Apple has been giving the Mac short shrift for years now and this puts the final nail in the coffin.

That's a big part that makes no sense to me, not to mention throwing so much legacy software down the drain. I think it reasonable that they could have an ARM as an extra processor, used as needed (and powered down when not needed), although that might need an IOMMU to help safely delegate limited hardware access to the ARM, not to mention a two-part OS. Alternatively, the ARM could be the main processor, with an Intel as the auxiliary; or that could migrate over time as appropriate.

 

[4look4rd]

My only concern with this is will this change still allow us to run Boot Camp and Parallels Desktop on ARM based Macs?

You'll probably be able to run Windows on ARM, which is fully featured but it runs x86 apps under a compatibility layer similar to Rosetta (when macs transitioned from PPC to Intel).