Apple is an existential threat to the PC

Unfortunately (?) not. I think you’ve compared to the older 7 GPU-unit A12x. The newer iPad Pros with all 8 enabled score higher still. Link.
For all intents and purposes it appears to be the same old iPad chip. Which is actually not a bad thing. Apple has really emphasized that the dev kit isn’t indicative of their actual future Mac silicon so the outlook is pretty nice. We’ll get a first indication once we see the next iPhone in a couple of months.

 

How do those new apple chips compare to say the CPU found in the next generation consoles?

 

PS5 is going to use an 8-core Zen 2 @ 3.5GHz, which is probably close to a Ryzen 7 3700X. On Geekbench 5, it has ~1200 single core performance and ~8000 multicore performance. So it's about 10% faster than A12Z on single thread performance and ~80% faster in multithread performance.

A comparison between A12Z iPad Pro and a 3700X iMac Pro

[EDIT] Note that the actual ARM based Mac is probably going to use something based on a future Apple CPU, probably A14, which is very likely to have better single core performance than A12Z (A13 used in iPhone 11 is already ~20% faster than A12Z on single thread performance in Geekbench 5).

 

Damn, that's pretty close for being what it is?

 

Comparing the A13 with the 3700X. Just look at that single core performance.

 

Where is the 3700x there?

 

It’s the ”iMac Pro”. (A so called ”Hackintosh” or ”FrankenMac”)

While extrapolating from Apples old mobile chips makes for some interesting speculation, it bears remembering that Apple may take their SoCs in a non-PC enthusiast direction. They may never make a ”power draw be damned” PC-tower lookalike using Apple Silicon. They are just as likely to keep absolute performance relatively restrained in order to keep power draw low and facilitate improved ergonomics in terms of size/weight/noise/battery life.

What is more disruptive in the market place, a laptop that has 50% higher performance or a one that has a 25h battery life and/or is completely fanless? Arguably, for most consumers, improved ergonomics beat improved performance.

That said, my personal belief is that the new Macs will be quite performant. Apple needs to give their customers the sense that shifting to their own SoCs represents a major step forward, and for their current and future Mac customers, performance is a significant factor, even if it isn’t the only or even most important one. Judging by their early statements, they are relatively happy with their current Mac product matrix, and initially prioritise changing underlying architecture rather than form factors. That may well be a misinterpretation though, and it’s eminently possible that their ambitions go further than doing-the-same-thing-only-better.

We just don’t know. The next couple of years are going to be interesting, but I have a feeling their Mac vision is far longer term than that.

 

Thanks intresting read. Prior this year i wasn't aware how fast mobile SoC's have become, both qualcomm and apple have really improved alot and can compete in the pc market, or atleast, close to that (not talking high end PC's with 2080 gpus).
If Apple is a thread to the 'PC'..... i don't know, maybe not a 'thread', but they could take over a share of AMD/intels place in the pc market. Windows would have to run on apple hardware then.

But as you say, no idea, i don't think anyone (even apple) does know. They want to take over the PC market or any relevant market offcourse, but

 

They could totally redefine the entire personal computing experience, just like they disrupted the entire smart phone market in 2007. Their mobile SoCs are basically fast enough for desktop class performance.

Imagine hooking your iPhone / iPad up to a monitor / dock, keyboard and mouse and it would switch from iOS to MacOS. That would be plenty fast for everyone except power users.

 

There are rumours that they have done exactly this. There is something appealing in having your personal unit of computing always available. It bears remembering though that a lot of people already use mobile platforms for all their everyday computing needs. Apple doesn't need MacOS to adress that.
Taking a larger view, iOS and MacOS will henceforth basically be separated by UI and filesystem handling, nothing says they intrinsically need to run on different underlying hardware. Turning that on its head though, if you do produce hardware in different form factors, it makes sense to use different underlying hardware to optimise for what the form factor allows. You build hardware to match and optimise for the ergonomics, rather than the other way around.

There is a huge scope in terms of possibilities. Presumably though, Apple want to build better Macs to adress the traditional computer market (the same market they have been cannibalising from the iOS end), so I imagine they want to increase their sales beyond what they have today and thus has to not only maintain their current customers in that niche, but also attract buyers from Windows. They are highly unlikely to gain much institutional market share, so they'll appeal to general private individuals and/or specifically adress lucrative niches. Initially, building something that can be convincingly be sold as "better computers" by direct comparison allows them to both maintain ASPs, and lure in new customers.

But the potential for out-of-the-box thinking is definitely there.

 

I think about this a bit differently.
I see it as extending the capabilities of the iPhone. You just have to add a bluetooth keyboard+mouse and connect to a display device, and you have something that allows you to edit photos and video for example on a screen that is more suited to the work. Or if you want to write the occasional longer text, or... the kind of stuff that justifies having a low- to mid-end PC around at all. Which, of course, is the bulk of the market. And everything you do is automatically backed up to the cloud.
In one fell swoop, you enable one and a half billion or so phones to adress a significant part of that portion of personal computing that really is better to do using a PC type device.
The iPhones are easily fast enough, leaked A14 benchmarks are better than any of my four private PCs, utterly crushing the laptops. Nevermind the capabilities of the image and video coprocessors.
The question is if running MacOS on the phone is the path to take (I’d love it). Apple are already moving iPadOS in this direction, the only difference being that the iPad screen sizes makes it unnecessary to connect to an external screen (even if you can do it). Direct handling of files is more a legacy/professional environment concern, although being able to futz around a bit with work when at home is also quite useful.

I have burned a lot of cycles professionally, but my personal computing needs are far more modest and would be quite well served by the new iPhone SoCs. Three out of four household PCs would be completely unnecessary if all I had to do was add a KB+M to my phone. And looking around me, that seems to be true for pretty much everyone who is not hooked on AAA gaming.

 

I think phones/tablets (not just apple ones) have already taken over the pc for most day to day home use. Its what im seeing around me atleast.

I have a 5s as a backup phone that still does most things OK, same for a xperia z2, both 2013/2014 phones

 

Beat this Apple!

 

hehe, so that's a Steam Machine

 

While some companies already tried the "docking" model and failed, the reasons why they failed are practically the same: they don't have a real desktop OS.
Now Apple is moving to converging iPadOS and macOS, and the app eco-system is likely to go forward with that, it'd be very interesting to see if Apple is willing to take this direction.
For example, if Apple made a simple dock with a HDMI/DP connector and a few USB ports for keyboard and mouse, and when you connect the dock to your iPhone, it goes into an iPadOS/macOS mode, it could replace a lot of home desktop PC. It might eat into some of the iMac markets, but I think it's going to be a great replacement for Mac minis (which, unfortunately, was neglected by Apple for some years). For millions of iPhone users, it's essentially an almost "free" new desktop computer, if they happen to have a monitor, a keyboard, and a mouse lying around. The dock can even function as an optional power adapter so you can charge your phone when you use it as a desktop computer at the same time.

Honestly I don't have high hope of Apple doing this, but I'd buy it without question if they decided to make one.

 

As you may be aware, iPadOS is being taken in this direction, the keyboard+mouse+display interface is already in place, in a functional but a bit primitive state. From there to iPhones is probably more or less the flipping of a software switch (+ marketing, Apple is only just starting to point out the possibility to the already converted). I’d prefer MacOS, because I’m just more comfortable getting hands on with the file system, and an OS tailored to those input devices in general.
As I mentioned, there is just something very appealing to having all your computing needs in your pocket, adaptable to whatever I/O mechanisms that suits your tasks. I’ve wanted that since my PalmPilots.

So if people already use cell phones to communicate, take and edit pictures and video, maintain calendars and schedules, take notes, get directions in cars, play games, handle their banking and do their taxes - what do they need PCs for at all?

Steve Jobs take on that was that traditional computers were like trucks - tools to get jobs done that are too heavy and/or impractical to with mobile devices (2 great minutes, where he also observes how this will make oldtimers uncomfortable. He was right.) And in this transition to Apple Silicon the company top brass have been very clear in their messaging. The first observation is that they all talk about performance, not just Johny Srouji. They not only seem confident in their competitiveness, they describe much improved performance as a prime reason why they transition at all. This is in line with their view of computers vs. mobile devices, if you build trucks and cars you want your trucks to offer greater capacity than your hatchbacks, otherwise what’s the point? The other thing that stood out to me was the absence of "enabling new designs and form factors". Not even mentioned once. Rather, they talk about matching silicon capabilities to existing form factors appearing positively conservative, for Apple.
While they will probably want to change things around a bit to separate old from new, enabling new designs was not a motivator, and thus not how they will sell their new systems. (Goodbye, Jony Ive. Maybe we can have ports on laptops again.)
Judging from their presentation of this transition, I have good hope that their new systems will pack a punch. It is also a good excuse to keep ASPs up. If adding 70mm2 of silicon die area allows a $400 higher price, that’s great for margins. And customers actually get something for their money. Considering what AMD delivers in their 7nm console SoCs, what Apple could deliver on 5nm or better might be quite impressive - and actually make their trucks capable of things their hatchbacks still can’t.

 

I am curious if this move might enable a better experience for the rumored Apple AR glasses though. The rumors I've heard say that most of the processing for these will be handled by an iPhone, presumably so that the glasses can stay thin and light, not scorch people's temples / ignite hair, etc... I'm not sure it really makes sense, but those duties could be taken over by ones Apple Silicon Mac when it is close - the much higher powered machines could preserve phone battery life and presumably provide higher fidelity to boot.

 

Possibly. That’s speculation about how one unreleased product may interact with another unreleased product though.
I haven’t heard any recent rumours about Apple using 802.11ay, unfortunately, which otherwise seems ideal for connecting wirelessly to AR/VR headsets.

 

Someone on Twitter claimed that Apple Silicon actually has a special "x86 memory model" mode for x86 emulation.



It's not unexpected, as x86's stronger memory model is a big problem when emulating x86 from an ISA with a weaker model (practically every RISC ISA out there has a weaker model).