The relationship between the iPad and the iPhone, performance-wise, has been hard to predict. I often point out that Apple is a company of annual patterns, often predictable. There’s not much of a pattern with regard to how the iPad and iPhone relate to one another.
Two years ago, when the original iPad Mini debuted, it was roughly a year behind its 9.7-inch sibling, the iPad 4. The original Mini had a non-retina display and A5 system-on-a-chip (SoC). In broad strokes, Apple took an iPad 2 and shrunk it to fit in a much smaller form factor. The iPad 4 had a retina display and an A6 SoC — and double the RAM (1 GB vs. 512 MB), a better camera, etc.
Last year, the new models were roughly all on par CPU/GPU-wise, with the A7 SoC running at about the same speed on all three devices: iPad Air, iPad Mini 2, and iPhone 5S. The Air had one small advantage over the other two devices: it was clocked at 1400 MHz instead of 1300 MHz, which gave it about a 5 percent advantage in CPU performance. The iPhone 5S had unique niceties, though, maintaining its clear position as the king of the iOS hill: a far superior camera and Touch ID, to name just two. The Air had better color gamut than the Mini, but I think it was very fair to say (which I did) that they were more or less the same iPad in two different sizes. The main thing you got when you paid the extra $100 for last year’s Air (versus the comparably-equipped Mini) was the size of the display.
This year, all previous patterns are busted.
The new iPad Mini 3 really just gets two things: Touch ID and a gold case option. Really, that’s it. Everything else about it remains unchanged.
The iPad Air 2, though, is entirely new. It’s a thorough refresh, that not only makes it a nice year-over-year improvement over last year’s iPad Air in just about every single regard, but arguably positions it above the iPhones 6 as the top-tier iOS device, period.
Let’s talk performance. The iPhones 6 still have just 1 GB of RAM. The iPad Air 2 has 2 GB. The iPhone’s A8 SoC has 2 billion transistors and two cores. The iPad Air 2’s A8X SoC has 3 billion transistors. According to Geekbench 3, Apple achieved this by going from two CPU cores to three. And the Geekbench benchmark results bear this out:
|Device||Single Core||Multi Core|
|iPad Air 2||1,813||4,539|
|iPad Air 1||1,472||2,664|
|11-inch MacBook Air (2011)||1,773||4,137|
|11-inch MacBook Air (2012)||2,560||5,170|
|13-inch MacBook Pro (2014)||3,518||7,438|
(The iPhone 6 and iPhone 6 Plus are so nearly identical in Geekbench results that I simply averaged the two together under “iPhones 6”.)
The Air 2 is noticeably faster than the iPhones 6 in single-core performance, but it’s simply in an altogether different ballpark in multi-core. I couldn’t get an answer from anyone at Apple regarding whether Geekbench is correct that it’s a three-core CPU,1 but the multi-core results certainly bear that out.
It is remarkable not only that the new iPad Air 2 is faster than the iPhones 6, but also that it’s faster than a three-year-old MacBook Air, and within shooting distance of a two-year-old MacBook Air. It’s more than half as fast as today’s top-of-the-line 13-inch MacBook Pro, especially in multi-core. (Let’s not get carried away regarding this apparent third core. Single-core performance is a better measure for most of the things typical consumers do on an iPad. But still.)
Geekbench is just a benchmark, but I think it’s a fair one for a general comparison of “how fast” machines are, including machines that run entirely different operating systems.
But Geekbench only measures CPU performance — year-over-year, the iPad Air 2 is even more impressive in terms of GPU performance. Apple claims, “The A8X chip has an astonishing 2.5 times the graphics performance of the A7 chip,” and from what I’ve seen, that’s true.
The iPad is no longer following in the wake of the iPhone, performance- and specs-wise. It’s forging ahead. With 2 GB of RAM, it’s a year ahead of the iPhone (we hope) in that department. Performance-wise it’s fast enough to replace a MacBook Air for many, many people. The demos that Apple chose for last week’s event — the Pixelmator image editor and Replay real-time video editor — emphasize that. Those are performance-heavy tasks, and the iPad Air 2 handled them with aplomb.
The other factor is Metal, Apple’s new low-level graphics API, which is (at least for now) iOS-only. Apple promotes it as being 10 times faster than OpenGL. For games and creative productivity apps that take advantage of Metal — and developers of both seem to be buying in — the iPad Air 2 might be on even footing today with the MacBook Air. It really is desktop-PC-level performance.
In short, I don’t think performance is any longer a reason to buy a MacBook Air instead of an iPad Air. The choice comes down to form factor and personal preference. This marks a turning point.2
The original iPad in 2010 didn’t even have a camera — not on the front, not on the back. Now, the iPad Air 2 has both a FaceTime camera and a rear-facing (“iSight”) camera that, to my eyes, produces images with quality somewhere between that of the iPhone 5 and 5S. It suffers in low light compared to the 5S, but in sunshine, it’s pretty close.
Apple didn’t anticipate people loving to use their iPads as cameras. I certainly didn’t either. But they do, and now Apple is embracing iPad photography. And for whatever the iPad Air 2 camera lacks optically, it makes up for many things through software. Again, I think the Pixelmator/Replay demos were carefully chosen — not only did they show off the Air 2’s computational performance, they exemplified the sort of apps you might want to use if you’re using the iPad as a camera. Shoot, edit, filter, publish and share — all from one device. You can do the same from an iPhone too, but other than the optical quality of the camera (where the iPhone 6 still has a noticeable edge), the iPad is the better machine for everything else. It’s faster, and its bigger display is better for composing new shots and examining/editing the shots you’ve already taken.
And then there’s the social aspect. The iPad camera has spawned an entirely new class of applications for teachers and coaches. Apple showed a clip of Coach’s Eye during the event last week, to name one very popular example. The iPad’s size enables a teacher and pupil, a coach and player, to share the display in a way that isn’t practical on a phone (even the 6 Plus).
When the iPad debuted in 2010, the primary question was what sort of things would we want to use them for instead of our iPhones and laptops. The answers vary by taste. Email, web browsing, watching video — a lot of the things we used to do on phones and laptops we now do on tablets.
Using the camera, in conjunction with smart software, as an instructional aid is something else: brand-new territory. Something only a tablet is suited for.
The iPad 3 was the first to go retina, and it was glorious to behold. But at 1.46 pounds, it was not so glorious to just plain hold. Not only was that 50 grams heavier than the iPad 2, it was 0.6 mm thicker. Apple products get thinner and lighter over time, not thicker and heavier. (The iPad 4, released just six months after the 3, was the same size and weight. It had better performances and switched the 30-pin adapter (gross, right?) for Lightning.)
Two years later, the iPad Air 2 is an entire half pound lighter than the iPad 3 was (0.98 pounds, down from 1.46) and 33 percent thinner (6.1 mm, down from 9.4 mm). A one-third reduction in thickness and weight in just two and a half years — with the same battery life, incredible performance increases (see above), and vastly improved cameras. The camera quality deserves special recognition, because, all things considered, the image quality of a camera tends to be inversely proportional to the distance between the lens and the sensor. (That’s why the iPhone 6 has a lens that juts out from the back of its frame.)
Yet the iPad Air 2 got thinner and its camera quality has increased. Part of that is that now that Apple is taking iPad photography more seriously, they’re using more expensive camera components. But part of it is due to things other than the camera lens and sensor. Face detection for auto-focus goes through the A8X’s image signal processor (presumably, the same ISP as on the iPhones 6), and the ISP plays a major role in noise reduction, burst mode, slo-mo, and more.
The end result is a markedly improved iPad, just in terms of it being an object you hold in your hands. It really does feel like we’re getting close to just holding a piece of glass. It’s very thin, very light, and very comfortable to hold. The improved display is a noticeable improvement over all previous iPads. Retina iPhone displays have been laminated to the glass touch screen ever since the first retina model (the iPhone 4, back in 2010). It really does feel like the difference between pixels-under-glass and pixels-on-glass. Now the iPad Air 2 offers the same thing, and it’s gorgeous. Even better, the iPad Air 2 one-ups the iPhone 6, with an anti-reflective coating. It’s quite noticeable, and very welcome. On a dark screen, it’s the difference between being able to see a reflection of my own face on the display, and being able only to see a silhouette of myself. I hope and expect this anti-reflective coating to spread to next year’s new iPhones and iPad Mini. The anti-reflective coating appearing on the iPad Air 2 first is another sign that the “new” iPhone is no longer the recipient of all new model-year improvements.
One other change is worth noting in the Air 2: Apple removed the “side switch” above the volume buttons. It was an on/off toggle that could be set to function as either a silent switch or a rotation lock (configurable in Settings). Now, it’s gone, and both those functions can only be controlled via the swipe-up-from-the-bottom-of-the-screen Control Center. I only ever used it as a silent switch, but I did use it. I’m curious what Apple’s rationale was for getting rid of it, because I thought it was useful.
Everything Apple is promoting about the Air 2 is true, both in terms of what you can objectively measure, and in terms of how it feels to use it. It’s thinner, lighter, faster, and has a better display and better camera. And, yes, Touch ID is great, especially if you’ve been using it for the last year on your iPhone.
I don’t think I’m going to buy one, though.
For the last two years, my day-to-day iPad has been a Mini. I like the Mini form factor so much that I switched to the original non-retina model in late 2012 even after having used the retina iPad 3 for six months or so. In terms of visual acuity, that was painful. In terms of hold-ability, though, it was a huge win. Last year I didn’t hesitate to stick with the Mini form factor once it went retina.
I spent a lot of time in this review comparing the new Air 2 to the iPad 3/4. I think that’s fair, because normal people aren’t supposed to even consider replacing their iPads on an annual basis. And from what we’re learning as the iPad era marches on, iPad users aren’t even upgrading them as often as they do their iPhones. They’re more like PCs, where people use them for several years. Anyone upgrading from an iPad 3/4 to an iPad Air 2 is going to be delighted. Anyone upgrading from an iPad 2 or original iPad is going to be amazed.
If you already have an iPad Air, it’s obviously a closer call. From the outside, Touch ID looks like the biggest change, but I’d actually rank it behind the improved display (lamination and anti-glare) and CPU/GPU performance in terms of how big a difference it makes in use. And the biggest change of all might be the reduction in weight and thickness. If you hold your iPad for long stretches of time, it really makes a difference.
But I still prefer the Mini form factor. I’m not saying it’s better in general — only that it’s better for me, personally. More than anything else, I mostly read on my iPad. When I do type on my iPad, I tend to do it iPhone-style with just my thumbs. For reading, the Air comes close to being a better iPad for me. After just four days of testing it, my iPad Mini already feels a little thick. But for typing, I’m far more comfortable with the Mini.
Which brings me to the new iPad Mini 3. Apple loaned me one of those to test alongside the Air 2. There’s really not much to review, though. Touch ID and the gold color option really are the only differences from the last year’s iPad Mini 2. Here’s what I wrote last year, in my review of the then-new iPad Air and iPad Mini 2:
So the iPad Air is an excellent year-over-year update over the iPad 4 — double the performance, and a serious reduction in size and weight. But the retina iPad Mini is an almost unbelievable year-over-year update — four times the performance, a retina display (which therefore means four times the pixels), and yet no appreciable difference in size or weight. This is the iPad Mini I expected to see next year, in 2014. But here it is today, in my hand.
Turns out I was right — we did get this year’s iPad Mini a year early. And now we still have it. But that’s OK. I think the sort of person who prefers the Mini form factor is less likely to be using their iPad in the ways that the iPad Air 2 is improved. (Anecdotally, most iPad photographers I see in the real world are using 9.7-inch iPads, not the Mini.) And the sort of iPad users who are pushing the performance limits of the platform are the sort of people who’ve preferred the 9.7-inch models all along. In short, I think the Mini really is more of a pure consumption device, and the Air is more of an alternative to a MacBook.
I’ve seen criticism that Apple now offers too many iPad models to choose from. The array of iPads for sale — three generations, two sizes, four storage tiers (16/32/64/128), and cellular-vs.Wi-Fi-only — is certainly not simple. But I don’t think it’s that tough for a would-be iPad buyer to decide. I’d say there are only four questions:
If you answer yes to question 3, you have to answer another question: Which carrier? But with Apple SIM, that’s no longer a long-term commitment unless you choose Verizon.
Question 4 is the tricky one, because you have to evaluate multiple factors, all of which cost additional money: performance, Touch ID, thinness/weight, and of course storage capacity.
It’s not so much that Apple has complicated the iPad lineup, as that they’ve expanded it downward into lower price points. They now have models ranging from $249 (the A5-based original iPad Mini, a.k.a. the zombie iPad) to $829 (the 128 GB cellular Air 2). At any given price point, there aren’t many decisions to make beyond color. I don’t know that that’s any harder a decision to make than buying a MacBook, especially once you factor in the various build-to-order upgrades that MacBooks offer.
On that last point, I’ll reiterate what I wrote a few weeks ago regarding the iPhone lineup. Apple should not be selling 16 GB iPads. The starting tier for typical consumers should be 32 GB. There’s just not enough usable space on a 16 GB iOS device to do the things Apple has worked so hard to make easy to do. High-def slo-mo video? Panoramic photos? Console-quality games? Those things all consume large amounts of space.
I heard from a few DF readers in education and the enterprise who said that their organizations buy 16 GB iPads and they all have plenty of free space to spare. It’d be a waste to force them to buy 32 GB models. I have no doubt that’s true. And I have no doubt there are millions of consumers for whom 16 GB is more than enough storage. But I don’t think it’s enough for the majority of typical iPad users, and that’s what matters.
I also understand the product marketing angle. That there are a lot of people who will look at the 16 GB models, see that they can get four times the storage for just $100 more, and buy the 64 GB model instead — when they would’ve bought the base model if it were 32 GB. I get it. There’s no doubt in my mind it’s good short-term business sense to go with a 16/64/128 lineup instead of 32/64/128. But Apple is not a short-term business. They’re a long-term business, built on a relationship of trust with repeat customers. 16 GB iPads work against the foundation of Apple’s brand, which is that they only make good products.
Apple has long used three-tier pricing structures within individual product categories. They often used to label them “Good”, “Better”, and “Best”. Now, with these 16 GB entry-level devices, it’s more like “Are you sure?”, “Better”, and “Best”. Fine, keep the 16 GB models around for expert business and education buyers who know that they really don’t need more storage space. But don’t put devices on the tables in Apple retail stores that you wouldn’t recommend as a good product and good value to typical customers.
If you’re on the fence about buying a 16 or 64 GB new iPad, especially the Air 2, I strongly encourage you to spring the extra $100 for the 64 GB. You’ll thank me later.
“Apple claims that the A8X has three billion transistors, compared to two billion on the iPhone 6’s A8. Is that because it includes a third core?”
Pause. “We’re not talking about that sort of detail regarding the A8X.”
“Would you call that an interesting question?”
Laughs. “You always ask interesting questions.” ↩︎
These CPU and GPU performance gains are all thanks to Apple’s own in-house chip design team. And Metal is all-Apple. They’re not sharing these chips or APIs with anyone else. It could well turn out to be a temporary historical blip, but at the moment, it looks like one company, Apple, is single-handedly pulling away from the entire rest of the industry in terms of computing performance and efficiency. ↩︎