The new iPads are the clue to the future of consumer and professional computing. The key to understanding this is to know how incredibly capable these thinnest of devices have become.
How powerful exactly? They’re now comfortably in the same park as MacBook Pros. Not just thinly powered MacBook Airs and MacBooks, but MacBook Pros: the powerhouse laptops used by content creators and other professionals. In fact Apple is claiming that the iPad Pros are more powerful than 9/10ths of current laptops. Let that sink in for a minute.
Of course there are still big issues that would deter professionals. IOS is not Mac OS by any stretch. With IOS there’s much less fine-grained control over your files. You can’t (yet) plug in external storage - although we remain hopeful that they newly provided USB C port might be more capable than it seems at first. You can’t even use a mouse, and there’s typically no trackpad on a typical iPad keyboard. I have low resolution fingers, and I need a better way to point on an iPad.
There’s a lot that still needs to change, but there’s increasingly little reason why it shouldn’t.
Up until now, I’ve been wondering whether Apple will eventually end its reliance on Intel CPUs in its MacBooks. It must be a constant source of irritation to Apple that it doesn’t use its own chips in its laptops. This must be worse than ever now that Apple’s A12X CPU is almost on a par with the Intel devices in the MacBooks.
There’s no physical or design-based reason why Apple can’t switch to using its own chips today in its laptops. When you attach a keyboard to a modern iPad it’s hard to tell that it’s not a laptop from a distance. There may be a myriad of smaller details but there’s nothing, in my view, that would stop this.
The biggest issue would be porting all the software that runs on the Intel chips across to ARM-based architecture. (Apple may design and manufacture its own A series chips, but they still use an ARM based core). This type of transformation has happened before, when Apple moved from IBM Power PC chips to Intel, around ten years ago, and it was quite a process for all concerned. It’s not something that anyone would relish - especially application developers, for whom this would represent an enormous and unwanted overhead. Applications written in, say C, can always be recompiled but this is not a one-stop shop, because of architectural, UI, API and even and "cultural" differences. Monolithic mature applications would need years of work and dozens of man-hours just to make them work again, never mind at an optimum level.
I’m beginning to think, though, that this is not an issue that Apple wants to force right now, and it may not have to.
There are a lot of big applications (that typically run on Macs and Windows) that are overdue for a complete architectural overhaul. Computer programs that have been around for a very long time can become bloated, not just with tacked-on features, but with fixes and work-arounds that somehow become assimilated into the body of the application. At some point, the program becomes a tangled, multi-layered mass of code that is hard to support and which is obviously not running at full efficiently. Sometimes it’s better simply to start again.
But if you’re going to do this, then you absolutely need to make sure that you’re going to target a platform that’s going to be around for a long time, and not one that might potentially have a limited future in front of it.
Can you see where this is heading?
If you were responsible for drawing the future path for your company’s application software, and you had to take a decision that would be valid for perhaps the next decade or more, what would you choose? ARM or Intel? It would be really hard to decide - because on the one hand, Intel is a platform you know extremely well and because this it might seem to be a safe choice. But on the other - very few people would deny that ARM is in the ascendancy. If it’s close to parity with Laptop chips now, where will it be in five or ten years time - as the Intel x86 architecture, universal though it almost is - runs out of steam? Honestly, if you try to imagine ten years ahead, it's not hard to think that most computers will be powered by ARM chips, and only a small minority of very high-end devices will still use x86 CPUs.
All of which is only part of the story, as GPUs and AI compute chips take over the really hard work. It's probably this more than anything else that - at least in the short term - make the idea of an ARM CPU at the centre of everyday or even content creators'computing a likely option.
As a software developer, it’s one of the hardest decisions you could possibly have to take. You’re effectively betting your company’s future on predicting a decade ahead in an industry where there are seismic-scale surprises round every corner. But there is another way, and I think it might already be happening, right under our noses. First. look at the rise of the iPad Pro.
For most purposes, a four year old iPad is as good as a new one. And indeed Apple is happy to sell older iPads cheaply because there's a demand for them. For content consumption, there's no need for blockbuster power. But for anything more than that, iPads need to keep pace with the times, and that means being able to deal with content creation and authoring, as well as mere playback. And - not insignificantly - some users just like using powerful stuff. The new iPads are so fast that it’s hard to imagine that they’re not being posited as general purpose computing devices - at least in some sense positioned to become the primary computer in a typical user's work regime.
Which brings us back to the question of whether or not Apple’s laptops and their iPads are set to converge.
The thing is: I don't think they're going to converge. At least, not in terms of what's inside them. I simply think that ARM, supported by AI and GPU compute, will take over.
So, how is this going to happen?
A clue might be in the imminent arrival of Adobe Photoshop for IOS. It now seems likely that Adobe is hedging its bets by making some of its flagship applications capable of working cross-hardware platform. This is not an impulsive move. The amount of work needed to port an application like Photoshop means that Adobe must have had a strong degree of conviction that IOS and ARM is going to be central to content creators in the future.
And as iPads get even more powerful, there will be less reason to work on the ageing PC architecture. And look at the advantages: iPads have phenomenal screens, they're natively touch sensitive, work for ages on a battery charge, and they're incredibly portable and lightweight.
We're not quite there yet, though. iPads have missing features. No mouse pointing. No user-accessible file system. No ability to work flexibly with external storage. These latter two deficiencies are going to have to be fixed, and the signs are that they might. The next iteration or two of IOS might hold the key to unlocking the full potential of that new USB C port. Has anyone done a teardown of an new iPad yet? I haven't seen one, so I don't know if there's a Thunderbolt chipset in there waiting to be liberated by that USB C connector. Even without that, there's no reason why external storage support shouldn't be added to the iPad Pro. As soon as a proper file system is supported in the operating system, that is.
So, that's what's likely in my view. It's a complicated set of circumstances, but I do think it is going to lead to ARM being the dominant platform for content creators in a short space of time, and for application developers, while it might be a headache to port their massive software packages across to ARM, it's the best possible time to rewrite them from the ground-up.
What will happen to x86 PCs? They're not going to go away. There's still a case for buying and using them. But they are going to become increasingly specialised, and, ultimately, perhaps marginalised. None of this takes account of anything that Intel might do to respond.
And what about Macs? Will they move to ARM? Or will people simply move towards ever more powerful iPads? I think that's exactly what will happen. It will do so under our noses. At some point, Mac users will be enticed by super-iPads. They'll have full capable ports and a genuinely usable file system. It might be that the availability of a file system is what ultimately marks the distinction between iPads and iPad pros. Effectively, at this point, Apple will let the market decide.
But wait! What about the Mac Pro? Surely the imminent (2019) arrival of this definitively non-mobile device completely scuppers all the above speculation? Maybe not. I think there are three possible outcomes.
- All of the above is largely wrong. Apple is going to maintain and continue to develop MacOS and the Mac Pro will be a "conventional" Intel box with expansion slots.
- It will be Intel based but highly modular in the sense of having external expansion via Thunderbolt or even PCIe.
- It will not use Intel at all, but will be based on multiple Apple ARM-based processors, and the OS will be a derivative of IOS with an accessible file system and the ability to access and work with a variety of external devices and modules.
If the third option turns out to be the case, then it would mean that Apple only has to support one hardware platform (Apple ARM-based processors and essentially the same architecture for both the iPad Pro and the Mac Pro - with obvious differences, of course, determined by the form-factor).
And it would only have to support one OS
You can see why Apple might want to do this.