How Apple could use 3x scaling for the iPhone 6 and iPad Pro

Mark Gurman at 9to5Mac first reported that Apple was testing upcoming iPhones with 1704 x 960 resolutions using 3x scaling. This was surprising to many, myself included. Why not choose 4x scaling and just quadruple the pixel count again?

In light of iOS 8, 1704 x 960 actually does make a lot of sense for the rumored 4.7" and 5.5" iPhones, as well as for a possible 13" iPad Pro.

Recently developer Steven Troughton-Smith cleverly discovered support for split-screen app views for the iPad in iOS 8. He even recorded a video demonstrating how two-finger gestures could trigger these 1/3, 1/2, and 2/3 split-screen views in the iOS Simulator.

Note that the 1/3 and 1/2 views use iPhone UI layouts, while the 2/3 view uses an iPad UI layout. With Apple's recent introduction of size classes (based on 320 point widths) for targeting new iOS display sizes, this all starts to make sense. (Also new in Interface Builder: 'Resizable iPhone' and 'Resizable iPad' simulators.) If the new iPhones are 1704 x 960, they are 50% greater in resolution width than the current Retina iPhones. Appropriately, the 1/2 iPad split view is also 50% wider than the 1/3 split.

Based on that hint, consider the following hypothetical displays:

iPhone 6:

  • 4.7" 1704 x 960, ~416 PPI
  • 5.5" 1704 x 960, ~356 PPI

iPad Pro:

  • ~13.0" 3072 x 1536, 2:1, 264 PPI
  • Landscape-only

This iPad Pro display size would have 50% greater resolution than the current Retina iPads. Using the same panel technology as the iPad Air (264 PPI) would yield an ~13.0" panel, extremely close to the rumored 12.9" display size for the iPad Pro.

While the iPhone 6 supposedly features 50% greater resolution along both axes, an iPad Pro would only need 50% greater resolution along the longer axis. (The widths are what matter, since portrait iOS apps are generally easier to stretch vertically; consider the iPhone 5 transition.)

As a result, the UI views developers would have to target for iPads effectively share size classes across devices:

Retina iPads running iOS 8:

  • 1/3 view: iPhone 5/5s
  • 1/2 view: iPhone 6
  • 2/3 view: Retina iPads in portrait orientation (doesn't scale perfectly but still works)
  • Full view: Retina iPads in landscape orientation

iPad Pro:

  • 1/3 view: iPhone 6
  • 1/2 view: Retina iPads in portrait orientation
  • 2/3 view: Retina iPads in landscape orientation
  • Full view: iPad Pro

Thus Apple would minimize the total number of display densities developers have to target. (Note to developers: use Auto Layout!)

Why make an iPad Pro with an ~13" display? I assume the purpose would be to add a landscape physical keyboard that is not too cramped. Support for split-screen app views would allow for an iPad with a keyboard that can display portrait iPad apps despite its inability to be held in portrait orientation. An iOS laptop with a 2:1 display would be a short device, but there would also be no need for a trackpad.

Typing speed is important. While information density is a significant factor, I believe most of the criticism that tablets are less productivity-enabling than laptops is simply due to the absence of a physical keyboard. Apple could effectively address both issues with an iPad Pro, should it deem the tradeoffs acceptable.



As has long been obvious, almost all of the above was wrong.

More Retina Macs

My previous post described why Apple might very well launch an 11.88” Retina MacBook Air this year with a 2732 x 1536 display.

Apple is obviously planning other Retina Macs. Are there any other economies of scale in its display supply chain that Apple could leverage to bring about these products?


First, some background. Anand Shimpi did an excellent job detailing the current hardware barriers to Single Stream Transport (SST) 60 Hz 4K monitors in his recent Mac Pro review. In short, Apple is waiting for future displays that include the necessary input controllers capable of refreshing SST 4K at 60 Hz before it enables Retina scaling modes for these monitors in OS X. Only then will Apple deliver a 4K Thunderbolt Display.

Additionally, it is extremely unlikely that the new Mac Pro’s GPUs are capable of driving SST 4K at 60 Hz, and the 2013 Retina MacBook Pros do not have the necessary driver support in OS X. An Apple 4K Thunderbolt Display and its associated Retina scaling modes might thus only be compatible with future Macs.

What about the display technology itself?

The current iMacs and Thunderbolt Displays feature 16:9 panels. For Apple, Retina resolutions require pixel quadrupling, and standard 3840 x 2160 4K is quadruple the pixel count of 1080p. Testing this resolution at 27” yields an interesting result:

  • 4K 3840 x 2160 iMac/Thunderbolt Display at 27” = 163 PPI, the same pixel density as the non-Retina iPhones and iPad mini.

I agree with Marco Arment that Apple will eventually release 27” 4K iMacs and Thunderbolt Displays. I have no idea when these displays will come out, but at least some volume of the required controllers are shipping in the first half of this year.

The possibility of sharing display technology with the original iPad mini lends further credence to the likelihood of Apple shipping 3840 x 2160 monitors using 27” displays. Their arrival by the end of the year certainly seems plausible.



Apple thankfully added 4K Retina display modes for the late 2013 Retina MacBook Pros and Mac Pro in OS X 10.9.3. A 4K Thunderbolt Display has yet to be announced, and consumers will likely have to wait until Macs are released with hardware support for SST 4K60. Unfortunately, it does not appear that the 2013 Mac Pro has the necessary hardware.

An 11.9” Retina MacBook Air

Taiwan’s United Daily News first published rumor of a 7.85” 1024 x 768 iPad in October of 2011. Andy Faust lent credence to the story with a convincing argument the following March. Apple eventually announced the 7.9” iPad mini in October to little surprise.

Similarly, rumors surfaced last October of a 12” Retina MacBook Air; DisplaySearch appears to have been the original source. Generally credible analyst Ming-Chi Kuo also voiced support for the 12” Retina MacBook Air theory.

Most rumors are false, but it doesn’t hurt to consider this somewhat specific claim.

The original speculated resolution of 2304 x 1440 made little sense. What resolution would a 12” Retina MacBook Air feature instead? The 11.6” MacBook Air currently comes with a 1366 x 768 panel. Pixel quadrupling this resolution would represent the easiest path towards a Retina MacBook Air in terms of meeting performance and power constraints, yielding a 2732 x 1536 display.

What size would this panel be? Assuming that the rumor of an approximately 12” Retina MacBook Air has some veracity, why not something close to it?

It turns out that an ~11.9” Retina MacBook Air with a 2732 x 1536 resolution happens to have the exact same pixel density as the 9.7” 2048 x 1536 Retina iPads: ~264 PPI. It would make sense for Apple to take advantage of the same IGZO display technology it has been utilizing for the iPad Air by cutting these panels to this larger size. I suspect a new 11.9” Air would be a redesign of the 11.6” model with smaller screen bezels.

It is surely expensive to source Retina panels for Macs, since their shipment volumes are far less than those for iOS devices. Apple would likely leverage any cost savings it possibly could to reach entry-level pricing. By utilizing its existing iPad Air display manufacturing and gradually replacing both MacBook Air models with just one Retina model, Apple would save a lot of money. OS X’s Retina scaling modes would also partially offset the lack of a 13.3” display option.

Recall that Apple made the discrete GPU-less 13.3” Retina MacBook Pro thinner last year, but not the 15.4” model. These facts would match up nicely with a three-tier laptop segmentation strategy. Eventually, Apple would sell only three Retina MacBooks ranging from thinnest to thickest, stratified by price and screen size: 11.9”, 13.3”, and 15.4”.

For all these reasons, I think it would make a great deal of sense for Apple to ship an 11.9” Retina MacBook Air with a 2732 x 1536 resolution later this year. It would be marketed as the 11.9” or 12” Air and run Intel’s Broadwell processors.

This new model would certainly represent one step towards John Siracusa’s goal for Apple to “introduce more, better Retina Macs” in 2014.



User shurcooL on Hacker News pointed out that the 2013 13.3" Retina MacBook Pro only caught up to the 15.4" Retina model when Apple made it thinner. (Both are now 0.71" thick.)

To clarify, this is what I meant by a "three-tier" segmentation:

11.9" Retina MacBook Air, possibly slightly thicker than the current Airs
13.3" Retina MacBook Pro
15.4" Retina MacBook Pro

I imagine the old MacBook Air design would stick around for a while at entry-level pricing.


Update 2

Since writing this post, I have come to a better understanding of the display supply chain. Ensuring quality supply is actually much more important than any potential cost savings. Sharing proven display technology across products is often necessary because of the difficulties of sourcing high quality panels in great volumes. These challenges have led to Apple's famous display struggles in recent years.


Obvious Update 3