Monthly Archives: June 2012

Design Tradeoffs in the Retina MacBook Pro

Earlier this week, Apple announced the much-anticipated release of the first Mac with a super-high-resolution “Retina” display. While some people hoped/expected it would be a 15″ version of the MacBook Air, what they released was positioned as the future of the MacBook Pro family. However, it was a much thinner and lighter MacBook Pro than previous models, delivering what many of the people hope for a 15″ MacBook Air hoped for.

Of course, Apple wouldn’t be Apple, and the Internet wouldn’t be the Internet, if there weren’t plenty of people complaining about what Apple ended up releasing. Since its release, any post-sale upgradability of the MacBook Air has been an afterthought. The RAM is soldered to the motherboard, and most recent models use a non-standard form-factor for the solid-state storage. The MacPro, on the other hand, has typically supported upgrades to both RAM and the hard disk. In fact, the Pros have tended to be more expandable than promised. Many models have supported more RAM than Apple specified, and many people have removed the optical drive in order to fit more internal storage.

The Retina MacBook Pro though, is like the Air. The RAM is soldered to the logic board and it uses of yet another non-standard SSD form-factor.  I’ll be honest, these things stick in my throat a bit too, but when it comes down to it, I wanted:

  1. A MacBook with a 15″ Retina display that is…
  2. Significantly thinner and, most importantly, lighter, than the current 15″ MBP while still having…
  3. Decent battery life.

Taken together, something has to give, because clearly the retina display must suck a lot of power. The machine has a more power-efficient CPU, and a significantly larger battery than earlier models, but still delivers “just” a “7hr” runtime with the Retina display.

We can start with the optical drive. Good riddance. That saves some weight and bulk. Eliminating some of the legacy ports, like Ethernet (now available as a dongle off a Thunderbird port” also allows the case to be thinner, saving some more weight, but potentially compromising rigidity.

Which brings us to the glued in battery. Gluing it in would seem to eliminate the need for a subframe to hold the cells. Gluing in the battery may also allow it to contribute to the rigidity of the assembled computer, reducing the need for some metal in the case, and saving more weight. I’m not too worried about the practicality of a user replacing the battery because Apple has been pretty good about replacing out of warranty batteries for free, and for those that don’t qualify for the free replacement, the charge has been a not-unreasonable $100.

Clearly that wasn’t enough for Apple, which brings us to the RAM. Clearly, they’ve eliminated a little weight and bulk by forgoing socketed RAM, but I don’t think that was the most important part. I notice that they’ve soldered the RAM to both sides of the logic board. I think the biggest benefit of eliminating socketed RAM is design flexibility. They didn’t just eliminate the SIMM slots and the SIMM boards, they eliminated the need to route all those traces to an easily accessible location.

Which brings us to the SSD. I’m ruling out the 2.5″ form-factor as requiring too many compromises to the rest of the system. As for mSATA, it appears that all the mSATA SSDs on the market are short little boards, and clearly don’t have room for 768 GB of flash chips Apple offers as an option. Which still leaves the question of why they didn’t use the SSD form factor used in the old Airs. I can’t answer that. Maybe they changed it just to be dicks, but I think there is some underlying reason, like performance and reliability. By doing so, they may have made things harder for the one or two 3rd parties who had products that fit Apple’s old proprietary form-factor, but it wasn’t exactly a big, commodified market that offered end-users significant savings over Apple’s prices.

And now, one final point. Taken together, this all means that Apple gets all the $ from all the people who want to upgrade from the base model. This may offend you, but consider that it might allow Apple to hit their financial targets while offering a lower base price. Lets face it, the base model, with 8GB of RAM and a speedy 256GB SSD is a solid, capable machine that will serve a long time. It may not hold all your media, but it will hold a lot of it.

So, that’s the way I see it. I plan to buy one soon. I’m not sure if I’ll spring for the $200 upgrade to 16GB of RAM. I’m sure it will add to the usable life of the machine, but its also 10% of the cost of my next upgrade.  One thing is clear to me, there aren’t going to be many PC vendors who can make a credible competitor to this machine any time soon.  For one thing, I suspect that Apple has locked up supplies of the displays for a while. For another, I’m not sure there are any PC vendors who are willing to go as far as Apple in optimizing the design of the device.

Apple, Cricket & Virgin. Hints of the future?

Before I could finish a long overdue post on Apple’s evolving relationship with mobile carriers, events overtook me. Since I started writing that post, Virgin Mobile and Cricket Wireless will soon be selling iPhones for use with their pre-paid plans. The pre-paid plans offer significantly lower monthly rates than traditional iPhone plans from AT&T and Verizon, particularly, but the hardware is significantly more expensive than the subsidized prices offered by those carriers. I find these developments interesting for a few reasons.

First, I it fits the pattern I described in my earlier post in that Apple is experimenting with more expensive phones in combination with cheaper, more flexible service.

Second, it is coming sooner than I anticipated, in ways I didn’t anticipate. I thought it would start with TMobile, but it is instead starting with Sprint, which owns Virgin Mobile, and Cricket, which makes heavy use of Sprint’s network. While I thought this would start with TMobile, it makes sense that it is starting with Sprint, since Sprint, like TMobile, is a second tier carrier which needs to be open to new business models in order to stay in the game at all.

I’m interested to see what comes next. I expect we’ll hear more by the time the next iPhone is released, likely by fall. I wouldn’t be surprised too if iOS 6, which will be announced at Apple’s World Wide Developer’s Conference next week, contains new and improved features that help marginalize the mobile carriers.

Apple’s Evolving Relationship with Mobile Carriers

I’ve been thinking about Apple’s relationship to mobile carriers, and the related question of whether their overall marketshare matters for a while. I’m going to take a little time now to outline the evolution of my thinking, and the conclusions I’ve drawn.

It started when I realized how Apple was going to use iCloud to tie OS X and iOS apps together to provide a seamless, pervasive computing environment, which I finally got around to writing about last fall. That led to the realization that although iCloud was designed to be as efficient as possible when moving data around between devices, it was still bringing Apple into conflict with mobile carriers.  Apple was moving towards a future where fast connectivity was always available at the same time the carriers were raising prices and eliminating unlimited data plans.

One of the big innovations of the original iPhone is that Apple “owned” the customer relationship. Yes, people paid AT&T for service, and many people bought their iPhones from AT&T, but Apple controlled the phone, Apple provided the support, and the software updates, and AT&T could not load up the iPhone with their logos or crummy apps. This was unprecedented, and it scared the carriers, and drove them into the arms of Android. Nevertheless, the carriers still held an important piece of territory: your phone number. Numbers are portable, but as long as people are making phone calls and using SMS, its hard to switch carriers often.

However, we’ve seen a rise in services that bypass the carriers. GoogleVoice gives you a carrier independent phone number for SMS and phone calls. Apple FaceTime can use an email address as an identifier, and now, so does Messages, their SMS substitute. Add to those other services, like Skype, Viber, Ping and TextFree. The carriers are slowly but surely being marginalized, but for the time being, they are still in a position to dictate prices, because hardware subsidies and the accompanying service contracts inhibit the formation of a truly competitive marketplace. It’s clear to me though, something is going to have to give.

I think Republic Wireless may be a sign of things to come. Republic offers mobile phone and data service without a contract for $19/month. To take advantage of the service, you need to spend $200 on phone with a special version of the Android operating system that routes voice calls over WiFi when available. When WiFi isn’t available, they use Sprint’s network. It isn’t hard to imagine Apple moving towards something similar.

In broad outline, I think Apple’s goal is to ensure that the mobile carriers are unable to pose a challenge to Apple’s vision for the products and services they want to offer their users. Ideally, they would disappear from the customer’s view all together, becoming suppliers to Apple. How Apple achieves this remains to be seen, but I think the ingredients are:

  • Apple has lots of capital, and they use that capital to gain a favored position with suppliers to ensure that they have a ready supply of key inputs, like flash memory,  high resolution screens, and bleeding edge manufacturing tooling, while depriving competitors of access.
  • Mobile carriers need a lot of capital.
  • Second-tier mobile carriers, like Sprint and T-mobile, risk going out of business if they can’t bring in enough revenue to do the network upgrades they need to stay competitive.  This is why Sprint sells service to carriers like Republic Wireless.
  • T-mobile doesn’t yet have the iPhone in the US.
  • Moore’s law can manifest itself in improvements to price, performance, and power-consumption. So far, much of the progress in the smartphone industry has been channeled towards improving performance. I think we may be reaching a point of diminishing returns, performance-wise.  This may push down pricing to the point that carrier-subsidies are less and less relevant.
These factors could play out a few different ways, but I’ll suggest an illustrative example:
  1. T-mobile starts selling the next iPhone.  Apple gives T-Mobile very favorable terms on payments, allowing T-mobile to invest in their infrastructure. In exchange, T-mobile iPhone customers get unlimited data.
  2. The 2013 iPhone comes at a lower price. Existing carriers are happy that they can reduce their subsidy, and more customers find an unlocked iPhone within reach.
  3. The next iPad can be used with any compatible carrier.
  4. Unlocked iPhones start showing up in the US that can be used on any carrier. The hardware already supports this; when an iPhone 4s is “born” it is capable of connecting to GSM networks like AT&T and TMobile, and CDMA networks like Sprint and Verizon, but when the phone is sold, these capabilities are limited.
  5. Apple starts offering subsidized phones with their own mobile service.

However, since I started writing this post, there have been some new developments, which I’ll post about separately.