Apple's third generation version of the iPad, presumably due in early 2012, will have to make substantial improvements in order to follow a strong but incremental second act.
In July of 2010, I wrote "The next generation iPad and what Apple needs to deliver." Based on information gleaned from updates in iOS and information coming out of the semiconductor industry at the time, I updated that article in November and polished the crystal ball -- in my usual purely speculative way -- of what I thought that iPad 2 might look like or the features it should contain.
The iPad 2 is now here. For the most part, much of what I thought would be in the new model did actually come to fruition, but I got a number of surprises and also some dissapointments as well. The iPad 2 is clearly a strong second act following the first device, but the improvements are still more or less incremental.
In the third version, Apple is going to have to up the ante considerably, because its competitors are going to have plenty of time in 2011 getting their respective offerings honed and improved for the next product cycle.
Let's go through last year's predictions about various anticipated features in the iPad 2 to find out where the hit and misses were, and to see if there is room for improvement in any of these areas that could make their way into iPad 3.
Firstly we can say that right off the bat, the new iPad will almost certainly have the same built-in gyroscope that the iPhone 4 has in addition to the accelerometer. This is a no-brainer addition, as it vastly improves responsiveness and precision control for games and also will permit the creation of even more immersive augmented-reality applications on the tablet.
Yup, the iPad 2 did in fact get a gyroscope, likely the same or similar part that's in the iPhone 4. So all future iPads are going to have gyroscopes. Will there be new applications on iPad that will really take advantage of it? Does the iPad 3 need a higher precision gyroscope part than what is shipping today? That remains to be seen, especially when iPhone 5 ships and we see what features it contains.
The second is the question of a front-facing camera and FaceTime. While FaceTime is undoubtedly one of the best features of the iPhone 4 and the iPod Touch 4, it may prove difficult to implement in the iPad without radically changing the existing hardware design and the behavior of FaceTime itself.
Not only did the iPad 2 get a front-facing camera, but it got a rear HD-capable camera as well. In my original piece, I discussed some of the challenges of using front-facing cameras on full-size tablets. As of yet, nobody has put Facetime into practice on an iPad 2, so I can't comment yet on how well the software translates to a larger mobile device.
However, I would expect that the specs on both the front and the rear cameras on iPad 3 to be improved. iPad 2 uses a VGA-capable camera on the front and a "HD" camera that can do 720p video in the rear. Competitors which are shipping with integrated cameras in 2011 include the Blackberry PlayBook, which sports a 3MP video camera in the front and a 5MP in the rear, and the Motorola XOOM, which is 2MP in the front and 5MP in the rear.
It is thought that the iPad 2 only uses a sub-1MP camera in the rear, which is good enough to shoot 720p, but not 1080p video. It also doesn't have a flash of any sort (such as the LEDs used on selected smartphones) and can only take sub-1MP still photos.
However, one must consider how Apple determines what features best fit in a device which is shipping at a base price point of $500.00, what the actual consumer use case scenarios are, and how they can purchase enough volumes of those parts in quantity to actually pull off that pricing.
So while competitors may ship "higher-spec" than iPad 2 today, they still will have trouble coming up with prices that are in line with what Apple is coming up with, based on the Bill of Materials (BOM) that determines the overall component cost that is hampered by the lack of these vendors' abilities to manipulate and control their respective supply chains in the same way Apple can.
Still, we can speculate that 3MP and 5MP camera components might be able to be secured in volume by Apple by end of CY 2011 to ship a device with upgraded parts in early 2012.
It should also be stated that the next-generation iPad should be brought to parity or exceed the memory capabilities as its flagship phone counterpart. The iPhone 4 has 512MB of main memory in its A4 Package on a Package (PoP), but the iPad currently only has 256MB, the same as the iPhone 3GS.
So we don't know yet how much memory the iPad 2 ships with on the updated A5 SoC barring a X-Ray examination of the PoP, but current industry scuttlebutt seems to suggest that Apple went with a conservative 512MB on the device, while also increasing the memory bandwidth.
I think that I can speak for many of us that I was disappointed that Apple didn't go for 1GB like its Android competitors, but I can certainly understand their reasons for wanting to keep costs down. In the iPad 3, I would expect flash prices to have fallen enough in 2011 that next year's device will have 1GB of RAM.
Central Processing Unit (CPU)
The current 1Ghz A4 System-on-a-Chip (SoC) processor in my opinion, interestingly enough, has been more than ample, but a dual-core and/or higher-clocked version of the chip with more integrated cache and improved integrated PowerVR graphics is probably in the works.
Apple delivered the goods with a dual-core A5, based on the ARM Cortex-A9 architecture, with substantially improved graphics performance. They didn't improve the clock speed, presumably to keep manufacturing design costs low and to keep battery performance at parity with the first model. An X-Ray examination of the PoP will actually determine whether or not the A5 is a totally new design, or simply two A4 cores tied together with an improved integrated GPU.
Room for improvement in the iPad 3 will be yet more graphics performance (particularly if the screen resolution is improved) such as moving to a PowerVR SGX6 GPU, increased clock speed and/or increased integrated L1 cache on the chip. The A5 has 64K of integrated cache (32K instruction + 32K data) like the A4. I'm not expecting a quad-core design in the next iPad on the A6 or A8, but I wouldn't put it out of the question that it exists in Apple's roadmap over the next two years.
The next is the issue of the display on the next-generation iPad, and this is where I think Apple could go down two different paths, depending on what market segment and what features they feel are more important.
It would certainly be possible to do a 9.7? display using the high-density 326 pixel-per-inch “Retina” technology, allowing for the display of true HD video at 720p or higher with an SXGA/UXGA resolution, but doing this might be hard for Apple to keep the entry-level version of the device at a $499.00 price point unless volume manufacturing of this technology improves significantly.
Clearly, costs of higher-resolution 9.7" screens did not go down significantly, so that's why iPad 2 shipped with the exact same spec screen as the previous model (although there is some indication they may have switched to a slightly thinner LG Philips part than previously.).
I don't expect that the iPad 3 will have the same resolution screen as an iPad 2 -- all indications are that the next model will almost certainly have a higher-res display, whether it is SXGA or UXGA or something even higher. In my earlier piece, I speculated that iPad 2 might have even used OLED or transfective-type displays, but this was a real reach on my part.
While I won't put some new technology out of the question for the iPad 3, it's possible that what we will see next year will be a higher res, but still a commodity IPS LCD part. OLED is still too expensive to source in quantity and obviously transflective technologies such as Pixel Qi and Mirasol aren't ready to use on tablets yet.
While a subject that is largely ignored in relation to the iPad, I also happen to think that the device needs a serious audio upgrade. Although I tend to use the device with headphones, I find the speakers to be anemic in terms of their overall volume and dynamic range — hopefully the next version will have nicer built-in speakers.
While I haven't yet experienced the quality of the speaker output of the iPad 2, I understand that they went from two smaller speakers that had monaural sound output in the iPad 1 to a single, but larger monaural speaker. With stereo devices coming out this year from Apple's competitors, I would expect iPad 3 to have stereo speakers.
I also think that given the video output capabilities of the latest generation of Android smartphones and Tablets, the next iPad should definitely have a way to output HDMI video, be it using an updated video output accessory (the current VGA kit is next to useless, even with Netflix support in the current software) or even with an actual integrated mini-HDMI port
Apple definitely delivered on this one. They created a new HDMI accessory that works on both iPad 1 and 2 and also included display mirroring capability on iPad 2, which is a big improvement in my book. The only thing I can think of to improve this on iPad 3 would be screen mirroring over Airplay to an Apple TV, but that might require some serious wireless network bandwidth.
The current model iPad only has a single-antenna Wireless-N transceiver and has a maximum network throughput of 65Mbps. For 1080p, you’d need at least two antennas and would want to double (or triple) that throughput, and have network performance much closer to what a 5Ghz Wireless-N laptop chipset can achieve.
We haven't done any throughput testing on the iPad 2 yet, but it would not surprise me if Apple stuck with a similar Wi-Fi transceiver on the new device in order to keep costs down and battery life where it needs to be. Obviously, I still think the iPad 3 would benefit greatly from a dual-antenna design that would permit full Wireless-N at 270Mbps and increased power for longer range signal.
Last year, it completely escaped me to talk about 3G. The 3G version of the iPad 2 apparently uses an integrated transceiver chip set that works on both CDMA and GSM 3G networks. However, unlike the Motorola XOOM and the BlackBerry PlayBook, the iPad 2 is not equipped to run on 4G this year. I'd expect that the iPad 3 will ship with models that support Verizon LTE as well as ATT's 4G network.
Lastly, there’s the issue of memory expansion, which seems to be a hot-button subject with the folks who responded to the original draft of this article over the summer — the iPad lacks an SD memory card slot, unlike like its Android competitors.
While I’d love to see such a slot implemented in the next design, unlike the other enhancements I spoke of above, it’s the one that Apple is least likely to implement, as they have a history of not allowing media to be directly side loaded onto their devices, and I don’t expect this policy to change.
And the policy didn't change. Apple didn't ship an SD slot on the iPad 2. I don't expect them to do it on the iPad 3 either. This is an area in which Apple's competitors will continue to differentiate and I think the company has drawn a line in the sand with, but you never know.
There are many other features which I would like to see in the iPad 3, but the ones I’ve described above are the ones that are most likely to make an appearance in 2012. What have I left out? Talk Back and Let Me Know.